Theory Seminar: Round&Round: An Improved Algorithm for 2-Dimensional Vector Bin Packing

Ariel Kulik (CISPA Helmholtz Center for Information Security)

Wednesday, 29.12.2021, 12:30

Taub 201 Taub Bld.

We study the 2-Dimensional Vector Bin Packing Problem (2VBP), a generalization of classic Bin Packing that is widely applicable in resource allocation and scheduling. In 2BVP we are given a set of items, where each item is associated with a two-dimensional volume vector. The objective is to partition the items into a minimal number of subsets (bins), such that the total volume of items in each subset is at most 1 in each dimension. We ...

[Full version]

[Full version]

Pixel Club: DPC: Unsupervised Deep Point Correspondence via Cross and Self Construction

Dvir Ginzburg (Tel-Aviv University)

Tuesday, 28.12.2021, 11:30

Zoom Lecture: https://technion.zoom.us/my/chaimbaskin

Deep Point Correspondence presents a new method for real-time non-rigid dense correspondence between point clouds based on structured shape construction. The method requires a fraction of the training data compared to previous techniques and presents better generalization capabilities. Until now, two main approaches have been suggested for the dense correspondence problem. The first is a spectral-based approach that obtains great results on synthetic datasets but requires mesh connectivity of the shapes and long inference processing ...

[Full version]

[Full version]

Deep Energy: Task Driven Training of Deep Neural Networks

The current gold standard in solving image processing and computer vision tasks is using supervised learning of deep neural networks (DNNs), requiring large-scale datasets of input-output pairs. In many scenarios in which the output is an image -- e.g., medical image analysis, image denoising, deblurring, super-resolution, dehazing, segmentation and optical flow estimation -- the collection of labeled image pairs for training is either time-consuming or limited to simple degradation models. Indeed, there is an increasing ...

[Full version]

[Full version]

Preparatory Meeting for the Employment Fair

Sunday, 26.12.2021, 19:00

Room 337 Taub Bld.

CS Student Baord invites you to a unique preparation evening to be delivered by 4 experienced engineers from the industry and the founders of the "חברמי" project - a non-profit social project, designed to help students and computer science graduates integrate into the high-tech industry. סלביק נימר - Principal Engineer at Microsoft גל דרוקר - Graduate of the Faculty and Tech Lead at Amazon אלדר סחייק - Software Engineer and Career Start Mentor ערן ליפשטיין ...

[Full version]

[Full version]

CS LECTURE: Theoretical Foundations for Emerging Multiprocessor Hardware

Naama Ben-David (VMware)

Sunday, 26.12.2021, 10:30

Room 012 Taub Bld (Learning Center Auditorium)

Due to the end of Moore’s law, hardware has been developing more rapidly in recent years than it has at any point since the early days of computing. These hardware developments are trending toward multiprocessor settings, which have the potential to deliver the speedups that CPU frequency scaling can no longer support. In this talk, I will discuss my work on building theoretical foundations for emerging multiprocessor technologies. I will focus on one line of ...

[Full version]

[Full version]

A Meeting with Potential Students for Technion CS Studies

Thursday, 23.12.2021, 09:00

Zoom Event: Registration

A meeting with potential students who are interested in studies at the Technion and the Faculty of Computer Science will be held online on Thursday, December 23, 2021. Details and registration

[Full version]

[Full version]

Theory Seminar: Secret Sharing, Slice Formulas, and Monotone Real Circuits

Benny Applebaum (Tel-Aviv University)

Wednesday, 22.12.2021, 12:30

Taub 201 Taub Bld.

A secret-sharing scheme allows to distribute a secret $s$ among $n$ parties such that only some predefined “authorized” sets of parties can reconstruct the secret $s$, and all other “unauthorized” sets learn nothing about $s$. For over 30 years, it was known that any (monotone) collection of authorized sets can be realized by a secret-sharing scheme whose shares are of size $2^{n-o(n)}$ and until recently no better scheme was known. In a recent breakthrough, Liu ...

[Full version]

[Full version]

End-to-End Referring Video Object Segmentation with Multimodal Transformers

The referring video object segmentation task (RVOS) involves segmentation of a text-referred object instance in the frames of a given video. Due to the complex nature of this multimodal task, which combines text reasoning, video understanding, instance segmentation and tracking, existing approaches typically rely on sophisticated pipelines in order to tackle it. In this work, we propose a simple Transformer-based approach to RVOS. Our framework, termed Multimodal Tracking Transformer (MTTR), models the RVOS task as ...

[Full version]

[Full version]

ceClub: Securing Legacy Communication Buses: Industrial Control Systems In-vehicle and In-aircraft Networks

Many important networking systems were designed decades ago, with a "closed environment" as a fundamental invariant: the networking infrastructure in a moving car, a flying aircraft, or a fenced power plant, were implicitly assumed to be isolated. As a result, the communication bus protocols were designed to function well despite natural phenomena such as noise, interference, radiation and so forth. No defenses against malicious adversaries were designed in. Once these isolated systems are connected to ...

[Full version]

[Full version]

ceClub: Theoretical Foundations for Emerging Multiprocessor Hardware

Naama Ben-David (VMware)

Wednesday, 22.12.2021, 11:30

Electrical Eng. Building 1061

Due to the end of Moore’s law, hardware has been developing more rapidly in recent years than it has at any point since the early days of computing. These hardware developments are trending toward multiprocessor settings, which have the potential to deliver the speedups that CPU frequency scaling can no longer support. In this talk, I will discuss my work on building theoretical foundations for emerging multiprocessor technologies. I will focus on one line of ...

[Full version]

[Full version]

CS LECTURE: Harnessing Scientific Literature for Boosting Discovery and Innovation

Tom Hope (University of Washington)

Wednesday, 22.12.2021, 10:30

Room 012 Taub Bld (Learning Center Auditorium)

With millions of scientific papers coming out every year, researchers are forced to allocate their attention to increasingly narrow areas. This creates isolated “research bubbles” that limit knowledge discovery and slow down scientific progress. Toward addressing this large-scale challenge for the future of science, my work explores new paradigms for helping scientists search and discover scholarly knowledge by developing novel approaches in information retrieval and AI/ML settings and models. In this talk, I will present ...

[Full version]

[Full version]

CS LECTURE: The Fine-Grained Complexity of Answering Database Queries

Nofar Carmeli (Ecole Normale Superieure Paris)

Tuesday, 21.12.2021, 10:30

Taub 301 Taub Bld.

We wish to identify the queries that can be solved with close to optimal time guarantees over relational databases. Computing all query answers requires at least linear time before the first answer (to read the input and determine the answer's existence), and then we must allow enough time to print all answers (which may be many). Thus, we aspire to achieve linear preprocessing time and constant or logarithmic time per answer. A known dichotomy classifies ...

[Full version]

[Full version]

CS LECTURE: Matching in Evolving Environments

David Wajc (Stanford University)

Monday, 20.12.2021, 10:30

Taub 601 Taub Bld.

Traditional computational models consider a one-shot computation, run on a single machine, with full knowledge of the input. In contrast, in many modern applications, the input is evolving, or is too large to be stored on any one machine, and is therefore partially unknown. These aspects of modern computation require decision-making under uncertainty. For example, in ride hailing applications, where drivers and riders arrive over time, the app needs to match users to each other ...

[Full version]

[Full version]

GoToNet: Fast Monocular Scene Exposure and Exploration

Autonomous scene exposure and exploration in localization- and communication-denied areas -- useful for finding targets in unknown scenes, mainly when direct maneuvering of the vehicle is impossible -- remains a challenging problem in computer navigation. In this work we propose a novel deep learning-based navigation approach that is able to solve this problem and demonstrate its ability in an even more complicated setup, i.e., when computational power is limited. Our method works directly with the ...

[Full version]

[Full version]

Quantum Candies and Their Applications

The field of quantum information is becoming more known to the general public. However, effectively demonstrating the concepts underneath quantum science and technology to the general public can be a challenging job. In this work, we present ``Quantum Candies'' (Qandies), a model for intuitively describing basic concepts in quantum information without the need for complex algebra or the concept of superpositions. We discuss several properties of Qandies, including their relation to quantum theory (Qubits), quantum ...

[Full version]

[Full version]

CS RESEARCH DAY 2021

Wednesday, 15.12.2021, 12:30

CS Taub Lobby

The 10th CS Research Day for graduate studies will be held on Wednesday, December 15, 2021 between 12:30-14:30, at the lobby of the CS Taub Building. Research Day events are opportunity for our graduate students to expose their researches using posters and presentations to CS faculty and all degrees students, Technion distinguished representatives and to high-ranking delegates from the hi-tech leading industry companies in Israel and abroad. The participating researches will be on various topics: ...

[Full version]

[Full version]

Theory Seminar: Extending Generalization Theory Towards Addressing Modern Challenges in Machine Learning

Shay Moran (Math, Technion)

Wednesday, 15.12.2021, 12:15

Taub 201 Taub Bld.

Recent years have witnessed tremendous progress in the field of Machine Learning (ML). However, many of the recent breakthroughs demonstrate phenomena that lack explanations, and sometimes even contradict conventional wisdom. One main reason for this is because classical ML theory adopts a worst-case perspective which seems too pessimistic to explain practical ML: in reality data is rarely worst-case, and experiments indicate that often much less data is needed than predicted by traditional theory. In this ...

[Full version]

[Full version]

CYBERDAY 2021

Monday, 13.12.2021, 09:00

Butler Auditorium (across CS Taub Building) , Technion

CYBERDAY 2021 event will be held on Monday, December 13, 2021 in the Technion on the topic of Operational Technology (OT) Security. The event is organized by Eli Biham, Yaron Gutmark, and the Technion Hiroshi Fujiwara Cyber Security Research Center. OT controls the modern industrial world, including pharmaceutics, power plants, hospitals, water supply, and almost any system that controls buildings or manufacturing, either critical or non-critical. Following Stuxnet and other attacks, the security of these ...

[Full version]

[Full version]

Recruitment Day by RAFAEL

Wednesday, 8.12.2021, 12:30

CS Taub Lobby

Rafael Representatives will arrive at the CS for demos and to present vacancies in the fields of software development, algorithms, cyber, hardware and electronics, on Wednesday, December 8, 2021, 12:30, on the entrance floor of the CS Taub Building. You are all invited.

[Full version]

[Full version]

Theory Seminar: Extremely efficient Distance Computations using distributed sparsity-aware Algorithms

Dean Leitersdorf (CS, Technion)

Wednesday, 8.12.2021, 12:30

Taub 201 Taub Bld.

Given a distributed network, represented by a graph of computation nodes connected by communication edges, a fundamental problem is computing distances between nodes. Our recent line of works show that while exactly computing all distances (All-Pairs-Shortest-Paths, APSP) in certain distributed settings currently has O(n^{1/3}) complexity, it is possible to find very good distance approximations even with an O(1) complexity. This talk will present a unified view of our various papers developing these interesting advances in ...

[Full version]

[Full version]

ceClub: From IP ID to Device ID and KASLR Bypass

Amit Klein (The Hebrew University of Jerusalem)

Wednesday, 8.12.2021, 11:30

Zoom Lecture: 95694696426

IPv4 headers include a 16-bit ID field. Our work examines the generation of this field in Windows, Linux and Android, and shows that the IP ID field enables remote servers to assign a unique ID to each device and thus be able to identify subsequent transmissions sent from that device. This identification works across all browsers and over network changes, including VPNs and browser privacy modes. In modern Linux and Android versions, this field leaks ...

[Full version]

[Full version]

COLLOQUIUM LECTURE - Cryptography from the Hardness of Kolmogorov Complexity

Rafael Pass (Cornell Tech)

Tuesday, 7.12.2021, 14:30

Room 337 Taub Bld.

Whether one-way functions (OWFs) exist is the most important outstanding problem in Cryptography. We will survey a recent thread of work (Liu-Pass, FOCS'20, Liu-Pass, STOC'21, Liu-Pass, Crypto'21) showing the equivalence of the existence of OWFs and (mild) average-case hardness of various problems related to time-bounded Kolmogorov complexity that date back to the 1960s. These results yield the first natural, and well-studied, computational problems characterizing the feasibility of the central private-key primitives and protocols in Cryptography. ...

[Full version]

[Full version]

Fast Distributed Algorithms via Sparsity Awareness

We show extremely efficient distributed algorithms for sparse matrix multiplication, distance computations (e.g. All-Pairs-Shortest-Paths, APSP), and subgraph existence problems. Our work identifies core observations regarding distributed computation and uses these to simultaneously tackle a variety of problems in several theoretical, distributed models. The central theme uniting our developments is designing sparsity-aware load balancing techniques and then applying them to problems on general, non-sparse, graphs.

[Full version]

[Full version]

Pixel Club: A Theoretical Analysis of Generalization in Graph Convolutional Neural Networks

In recent years, the need to accommodate non-Euclidean structures in data science has brought a boom in deep learning methods on graphs, leading to many practical applications with commercial impact. In this talk we will review the mathematical foundations of the generalization capabilities of graph convolutional neuralnetworks (GNNs). We will focus mainly on spectral GNNs, where convolution is defined as element-wise multiplication in the frequency domain of the graph. In machine learningsettings where the dataset ...

[Full version]

[Full version]

Workshop on Resume Upgrade

Wednesday, 1.12.2021, 08:30

Zoom Event: Registration

You are invited to a workshop on resume upgrade that will take place in a one-on-one conversation with a recruiter from a leading high-tech company who will help you improve your "business card" to the world of employment and job interviews in details such as: - How to be outstanding - How to deal with inexperience or with a low average - Use of keywords ("buzz") in search engines And more. After registering (by November ...

[Full version]

[Full version]

Workshop on Exposure to Industry Jobs: Web Development

Wednesday, 24.11.2021, 18:30

Zoom Event: Registration

CS invites you to a workshop on exposure to industry jobs in Web Development, which will discuss key words (buzz words) and various titles as well as the technologies and products it develops, will provide a glimpse of a typical workday, career development options and professional horizon, and will give a lecture by Yehonatan Lusky, CS Master graduate and CTO in Pikoya. The workshop will be held on Monday, November 24, 2021, 18:30 and a ...

[Full version]

[Full version]

Recruitment Day by Mobileye

Wednesday, 24.11.2021, 12:30

CS Taub Lobby

Mobileye representatives will come to the faculty to present their recruitment processes and vacancies, on Wednesday, November 24, 2021, 12:30, in the Taub Building Lobby, and a lecture on: Accelerating Autonomous Driving by Arie Tal, Principal Engineer Manager, will be held at 13:30 in the class 3 in the lobby - Please register in advance for the lecture. You are all invited.

[Full version]

[Full version]

Introductory Meeting with Researchers in the CS Theory Group

Wednesday, 24.11.2021, 12:15

CS Taub Balcony

You are invited to an introductory meeting with researchers in the CS Theory Group who will talk about structures in distributed networks, calculations on confidential information, correctness of calculations, the majority decision at trial and frequencies for radio stations. The event will take place on Wednesday, November 24, 2021 at 12:15, on the CS Taub Balcony (2nd floor). You are all invited!

[Full version]

[Full version]

Improving real data results, using Sim2Real models with GANs

Understating and controlling generative models' latent space is a complex task. In this work, we propose a novel method to learn the behavior of any specific attribute in an existing GAN's latent space and edit real data samples accordingly. We perform Sim2Real learning, relying on only three synthetic samples from two classes per attribute, allowing an unlimited amount of different precise edits. We present an AutoEncoder based model which learns both the essence of a ...

[Full version]

[Full version]

ceClub: Fast and Efficient Genomics Compute with the First Available Processing in Memory (PIM) on DRAM

Fast and efficient genomics compute with the first available Processing In memory (PIM) on DRAM Processing In memory (PIM), as UPMEM developed it - programmable processors integrated on DRAM die and orchestrated through the host processor -, are now widely accepted, and promoted as the next evolution in compute architecture. Memory vendors, JEDEC, now processor vendors, users, IT labs… are now exposing PIM in their plans. PIM’s benefits are based on structural considerations about the ...

[Full version]

[Full version]

Recruitment Day by DELL Technologies

Wednesday, 17.11.2021, 17:30

Zoom Event: Registration

You are invited to an online Recruitment Day with DELL Technologies engineers who will speak at an expert panel on "How to Get to the Top in Technological Innovation Development". The event will take place on Wednesday, November 17, 2021, 17:30 via Zoom. Please register in advance.

[Full version]

[Full version]

Theory Seminar: On Parameterized Analysis and the Disjoint Paths Problem

Meirav Zehavi (Ben-Gurion University)

Wednesday, 17.11.2021, 12:30

Taub 201 Taub Bld.

Parameterized Analysis leads both to a deeper understanding of intractability results and to practical solutions for many NP-hard problems. Informally speaking, Parameterized Analysis is a mathematical paradigm to answer the following fundamental question: What makes an NP-hard problem hard? Specifically, how do different parameters (being formal quantifications of structure) of an NP-hard problem relate to its inherent difficulty? Can we exploit these relations algorithmically, and to what extent? Over the past three decades, Parameterized Analysis ...

[Full version]

[Full version]

How to Improve or Attack Deep Uncertainty Estimation Performance

Deep neural networks (DNNs) must be able to estimate the uncertainty of their predictions when deployed for risk-sensitive tasks. In the first part of this talk, we present a comprehensive study that evaluates the uncertainty performance of 484 deep ImageNet classification models. We identify numerous and previously unknown factors that affect uncertainty estimation. We find that distillation based training regimes consistently yield better uncertainty estimations than other training schemes such as vanilla training, pretraining on ...

[Full version]

[Full version]

Complexity Measures on the Symmetric Group and Beyond

We study complexity measures of Boolean functions on the symmetric group and other domains. This generalizes classical work on functions over the Boolean cube. Additionally, we construct efficient circuits for low sensitivity functions. Using our theory we give an alternate proof for the characterization of extremal t-setwise-intersecting families of permutations. Joint work with Yuval Filmus, Noam Lifshitz, Nathan Lindzey and Marc Vinyals

[Full version]

[Full version]

Workshop on Exposure to Industry Jobs: Cyber Security

Monday, 15.11.2021, 18:30

Zoom Event: Registration

CS invites you to a workshop on exposure to industry jobs in cyber security, which will discuss key words (buzz words) and various titles as well as the technologies and products it develops, will provide a glimpse of a typical workday, career development options and professional horizon, and will give a lecture by Yehonatan Lusky, CS Master student and Second in the faculty and Head of Research Team at Intel. The workshop will be held ...

[Full version]

[Full version]

Using Fictitious Class Representations to Boost Discriminative Zero-Shot Learners

Humans do not learn all the classes they encounter all at once, rather they learn them gradually. Moreover, they can learn new classes with little or no examples. This has inspired new branches of machine learning research, such as zero-shot learning and life-long learning, which aim to replicate this ability in machines. In zero-shot learning, the model is required to recognize classes it had not previously seen in training data. This is usually achieved by ...

[Full version]

[Full version]

Codes over Graphs

In this work, we present a new class of codes, called codes over graphs. Under this paradigm, the information is stored on the edges of undirected or directed complete graphs, and a code over graphs is a set of graphs. A node failure is an event where all edges in the neighbourhood of the erased node have been erased. We say that a code over graphs can tolerate ρ node failures if it can correct ...

[Full version]

[Full version]

Beware of Geeks Bearing Gifts: Attacking Smart Homes

The smart home industry is increasingly growing and is expected to reach billions of homes around the world in the near future. Apple allows its users to easily and securely control their smart homes using its own proprietary protocol called HAP. Although HAP is considered secure, it attracted several attacks. A specially interesting one allows a remote attacker to steal the user’s home Wi-Fi password. Although this attack has a great impact on the home ...

[Full version]

[Full version]

Theory Seminar: Settling the Complexity of Nash equilibrium in Congestion Games

Yakov Babichenko (IE, Technion)

Wednesday, 10.11.2021, 12:30

Taub 201 Taub Bld.

We consider (i) the problem of finding a (possibly mixed) Nash equilibrium in congestion games, and (ii) the problem of finding an (exponential precision) fixed point of the gradient descent dynamics of a smooth function f:[0,1]^n -> R We prove that these problems are equivalent. Our result holds for various explicit descriptions of f, ranging from (almost general) arithmetic circuits, to degree-5 polynomials. By a very recent result of [Fearnley, Goldberg, Hollender, Savani ’21] this ...

[Full version]

[Full version]

Advanced UXSS Analysis

There are many types of security vulnerabilities and exploits that utilize them, and most of them are well studied. Yet, a family of severe security exploits called Universal Cross-Site Scripting (UXSS) has been hardly explored and the foundation required to study them has not been formulated. In this thesis, we focus on this family of exploits. A UXSS exploit enables the attacker to execute a controlled script in the context of any cross-origin service. UXSS ...

[Full version]

[Full version]

Auctions with Interdependence and SOS: Improved Approximation

Interdependent values make basic auction design tasks -- in particular maximizing welfare truthfully in single-item auctions -- quite challenging. Eden et al. recently established that if the bidders valuation functions are submodular over their signals (a.k.a. SOS), a truthful 4-approximation to the optimal welfare exists. We show existence of a mechanism that is truthful and achieves a tight 2-approximation to the optimal welfare when signals are binary. Our mechanism is randomized and assigns bidders only ...

[Full version]

[Full version]

Theory Seminar: On Edit Distance Oracles and Planar Distance Oracles

Oren Weimann (Haifa University)

Wednesday, 3.11.2021, 12:30

Taub 201 Taub Bld.

(1) Edit distance oracles: preprocess two strings S and T of total length n into a data structure that can quickly report the optimal edit distance between any substring of S and any substring of T. I will describe a data structure with query time Õ(1), and construction time and space n^(2+o(1)). This is optimal up to subpolynomial factors since computing just the edit distance between S and T requires quadratic time (assuming the strong ...

[Full version]

[Full version]

How Do You Turn a Degree into a Career?

Tuesday, 2.11.2021, 18:30

Zoom Lecture: Registration

How Do You Turn a Degree into a Career?You are invited to a lecture and conversation with Dr. Jonathan Yaniv, CS graduate and head of the research group at YOTPO, which will deal with the questions: Is it worth working during the degree? How to integrate into the industry, especially now in the Corona times? How do you find the right job for you? How to build a career path? Working in a large company ...

[Full version]

[Full version]

DARLING: Data-Aware Load Shedding in Complex Event Processing Systems

Complex event processing (CEP) is widely employed to detect user-defined combinations, or patterns, of events in massive streams of incoming data. Numerous applications such as healthcare, fraud detection, and more, use CEP technologies to capture critical alerts, threats, or vital notifications. This requires that the technology meet real-time detection constraints. Multiple optimization techniques have been developed to minimize the processing time for CEP, including parallelization techniques, pattern rewriting, and more. However, these techniques may not ...

[Full version]

[Full version]

Recruitment Day by NVIDIA

Wednesday, 27.10.2021, 17:30

Teams Event: shorturl.at/hnAST

CS students are invited to Recruitment Day by NVIDIA to learn about its activity, to hear lectures on Software, Firmware, Architectura, Chip Design and to meet its engineers who will answer you questions and tell you about their employment options. The event will be held via Teams on Wednesday, October 27, 2021, between 17:20-19:00 and requires pre-registration. During the event, a logic puzzle bearing a cash prize for a vacation in Israel or abroad will ...

[Full version]

[Full version]

Reducing Supervision in Visual Recognition Tasks

While deep neural networks (DNNs) have shown tremendous success across various computer vision tasks, including image classification, object detection, and semantic segmentation, requirements for a large number of high-quality labels obstruct the adoption of DNNs in real-life problems. Lately, researchers have proposed multiple approaches for reducing requirements to the amount or quality of these labels or even working in a fully unsupervised way. In a series of works, we study different approaches to supervision reduction ...

[Full version]

[Full version]

The Complexity of the Shapley Value for Path Queries over Graphs

A path query extracts from the input graph the pairs of vertices that constitute the endpoints of matching paths, that is, paths such that the word obtained from the edge labels belongs to a specified language. We study the computational complexity of measuring the contribution of edges and vertices to an answer of a path query. For that, we adopt the traditional Shapley value from cooperative game theory. This value has recently been suggested and ...

[Full version]

[Full version]

CS Orientation Day 2021-22

Thursday, 21.10.2021, 10:00

CS Taub Lobby and Taub Auditorium 1

CS 2021-22 Orientation Day for new students will be held on Tuesday, October 21, 2021, and will begin at 10:00 with a Technion meeting at the Kellner Amphitheater where the Senior Vice President, the Dean for Undergraduate Studies and the Students Dean and Chairman of the Technion Student Association will speak to the new students, and between 11:00-14:00 there will be a gathering at the Computer Science Taub Building, and a meeting at the Taub ...

[Full version]

[Full version]

Access Patterns and Adaptive Caching

Although caching is a well-studied topic and a widely used approach, modern software cache systems still struggle to optimize their decisions for many types of environments and workloads. This talk will briefly present four of our studies trying to tackle several on-going caching challenges. The first study reexamines the FIFO vs. LRU battle, showing that modern cache systems can often benefit from the simpler FIFO policy. The second study presents an adaptivity mechanism for software ...

[Full version]

[Full version]

Geometry In Numerical Algorithms

A common paradigm in engineering consists of problem modeling followed by numerical optimization. Over the years, a chasm has formed between the two stages: Models are becoming more and more complicated in order to address data irregularities while numerical optimization is being delegated to an external solver, usually not designed to handle the specific problem at hand. In this thesis I focus on problems in which there exist some underlying geometric or topological structure. Such ...

[Full version]

[Full version]

Complex Pattern Mining

Mining complex patterns from large data sets has attracted much attention in the last few decades. A plethora of methods and algorithms have been designed for mining a variety of patterns, ranging from simple association rules and frequent itemsets to advanced graph-based structures. However, as modern applications grow dramatically more sophisticated and operate on highly multidimensional and increasingly complex data, they introduce the demand for mining even more expressive and convoluted patterns unsupported by the ...

[Full version]

[Full version]

Poisson Denoising of Images Using Deep Neural Networks Inspired by Classical Dictionary Based Algorithms

The removal of noise from an image is an important and fundamental task in image processing. The statistical distribution of this noise is dependent on the measuring technique and the nature of the captured image and should be considered when tackling the denoising problem at hand. In some applications, such as in night vision, astronomy and fluorescence microscopy, the images are acquired under low light conditions and the image sensor counts a small number of ...

[Full version]

[Full version]

New Tools for Instance-Wise Predictive Uncertainty Estimation in Regression Problems

Learning algorithms are increasingly prevalent within consequential real-world systems, where reliability is an essential consideration: confidently deploying learning algorithms requires more than high prediction accuracy in controlled testbeds. In this talk, we will introduce recent advancements in quantile regression—a general technique for assessing the prediction uncertainty in regression problems. In the first part, we will modify quantile regression and introduce a novel loss function that drives the predictive model (e.g., a deep net) to accurately ...

[Full version]

[Full version]

Ranking and Trading Execution of Mean-reverting Portfolios

Statistical Arbitrage is one of the pillars of quantitative trading, and has long been used by hedge funds. Historically, statistical arbitrage evolved out of the simpler pairs trade strategy, in which stocks are put into pairs (a portfolio of two stocks) by fundamental or market-based similarities. When one stock in a pair outperforms the other, the under performing stock is bought long and the outperforming stock is sold short with the expectation that under performing ...

[Full version]

[Full version]

Online Partially Observable Markov Decision Process Planning via Simplification

Partially Observable Markov Decision Process (POMDPs) are notoriously hard to solve. In this work we consider online planning in partially observable domains. Solving the corresponding POMDP problem is a very challenging task, particularly in an online setting. Our key contribution is a novel algorithmic approach, Simplified Information Theoretic Belief Space Planning (SITH-BSP), which aims to speed up POMDP planning considering belief-dependent rewards, without compromising on the solution's accuracy. We do so by mathematically relating the ...

[Full version]

[Full version]

Digital Gimbal: End-to-end Image Stabilization with Learnable Exposure Times

Mechanical image stabilization using actuated gimbals enables capturing long-exposure shots without suffering from blur due to camera motion. These devices can be externally attached to any camera with no need for specialized optics, making them the most common stabilization solution; however, they are often physically cumbersome and require high amounts of power, limiting their widespread use. In particular, an alternative solution for light airborne imaging systems, which are inherently prone to motion blur due to ...

[Full version]

[Full version]

Dynamicity and Multi-commodity in Networks

The world is dynamic and changes over time, thus any optimization problem used to model real life problems must address this dynamic nature, taking into account the cost of changes to a solution over time. The multistage model was introduced with this goal in mind. In this model we are given a series of instances of an optimization problem, corresponding to different times, and a solution is provided for each instance. The strive for obtaining ...

[Full version]

[Full version]

Towards Understanding the Hardness of Multi-Agent Path Finding

The problem of Multi-Agent Path Finding (MAPF) calls for finding a set of conflict-free paths for a fleet of agents operating in a given environment. Arguably, the state-of-the-art approach to computing optimal solutions is Conflict-Based Search (CBS). In this work we revisit the complexity analysis of CBS to provide tighter bounds on the algorithm's run-time in the worst-case. Our analysis paves the way to better pinpoint the parameters that govern (in the worst case) the ...

[Full version]

[Full version]

Virtual Workshop on Machine Learning & Hardware Security

Wednesday, 1.9.2021, 12:00

Zoom Event: Registration

Objective: Exchange knowledge and research ideas in the area of using machine learning to improve evaluation and reverse engineering techniques. We hope to: - Create a community of researchers in these areas - To expose the researchers to new infrastructures, techniques and research topics - To establish a common ground for new collaborations and funding opportunities Registration is free but required - meeting details are shared with only registered participants by email, and in case ...

[Full version]

[Full version]

Offline and Online Algorithms for SSD Management

Solid state drives (SSDs) have gained a central role in the infrastructure of large-scale datacenters, as well as in commodity servers and personal devices. The main limitation of flash media is its inability to support update-in-place: after data has been written to a physical location, it has to be erased before new data can be written to it. Moreover, SSDs support read and write operations in granularity of pages, while erasures are performed on entire ...

[Full version]

[Full version]

Better Approximations for Bin Packing with Clique-graph Conflicts

We study the following variant of the classic bin packing problem.Given a set of items of various sizes, partitioned into groups, find a packing of the items in a minimum number of identical (unit size) bins, such that no two items of the same group are assigned to the same bin. This problem, known as bin packing with clique-graph conflicts, has natural applications in storing file replicas, security in cloud computing and signal distribution. Our ...

[Full version]

[Full version]

Langevin Dynamics in Image Restoration

Inverse problems in image processing refer to a family of problems in which we aim to recover an original signal given degraded measurements of it. Various techniques and algorithms have been suggested for general inverse problems, with a special emphasis dedicated to the most prominent example -- image denoising. Recent deep neural network approaches for these tasks focus on minimizing the mean squared error (MSE) between the original and the reconstructed signals. However, in moderate ...

[Full version]

[Full version]

Efficient Distributed Construction of Small k-Dominating Sets

We improve the message efficiency of the time-efficient construction of a "small" (i.e. universaly optimal) k-dominating set (k-DS) under the Distributed CONGEST model. This task was suggested by Kutten and Peleg as a useful primitive in constructing other time-efficient algorithms such as a minimum spanning tree. It is also useful for constructing other local (i.e. sub-diameter time) algorithms such as partitioning the network into clusters (each a rooted tree) of diameter k. We first address ...

[Full version]

[Full version]

Learning to Log with Control Flow Graph

Despite significant progress in software testing and verification, some undesired behaviors inevitably make their way to production. It is therefore common practice to interleave logging operations into modern software. Logging operations store information about the program's execution to help debugging and diagnosing problems. Usually, the programmer decides what parts of the program's state to log. This work aims to automatically complete logging operations in a given program based on learning from logging operations in other ...

[Full version]

[Full version]

Concurrent Data Structures for Non-Volatile Memory

Michal Friedman

Tuesday, 3.8.2021, 14:00

Room 601 Taub Bld.

With the recent launch of the Intel Optane memory platform, non-volatile main memory in the form of fast, dense, byte-addressable non-volatile memory has now become available. Nevertheless, designing crash-resilient algorithms and data structures is complex and error-prone, especially when caches and machine registers are still volatile and the data residing in memory after a crash might not reflect a consistent view of the program state. This talk will present different approaches and transformations that adds ...

[Full version]

[Full version]

On the Recursive Structure of Multigrid Cycles

A new fixed (non-adaptive) recursive scheme for multigrid algorithms is introduced. Governed by a positive parameter $\kappa$ called the cycle counter, this scheme generates a family of multigrid cycles dubbed $\kappa$-cycles. The well-known $V$-cycle, $F$-cycle, and $W$-cycle are shown to be particular members of this rich $\kappa$-cycle family, which satisfies the property that the total number of recursive calls in a single cycle is a polynomial of degree $\kappa$ in the number of levels of ...

[Full version]

[Full version]

Debiasing Methods in Natural Language Understanding Make Bias More Accessible

Model robustness to bias is often determined by the generalization on carefully designed out-of-distribution datasets. Recent debiasing methods in natural language understanding (NLU) improve performance on such datasets by pressuring models into making unbiased predictions. An underlying assumption behind such methods is that this also leads to the discovery of more robust features in the model’s inner representations. We propose a general probing-based framework that allows for post-hoc interpretation of biases in language models, and ...

[Full version]

[Full version]

Inferring Mitochondrial and Cytosolic Metabolism by Coupling Isotope Tracing and Deconvolution

The inability to inspect metabolic activities within distinct subcellular compartments has been a major barrier to our understanding of eukaryotic cell metabolism. Previous work addressed this challenge by analyzing metabolism in isolated organelles, which grossly bias metabolic activity. Here, we developed a method for inferring physiological metabolic fluxes and metabolite concentrations in mitochondria and cytosol based on isotope tracing experiments performed with intact cells. This is made possible by computational deconvolution of metabolite isotopic labeling ...

[Full version]

[Full version]

Generative Models: Affecting Current Practice with Traditional Methods

Inverse problems in the field of signal processing refer to the estimation of a (clean) signal when given corrupted or partial measurements of it. In this research thesis, we focus on solving such problems, using both the traditional sparse representation model and the more recent deep neural networks approach. In a series of papers, we show how one could utilize the mathematically well-understood results of the former, to improve the common practice of the latter, ...

[Full version]

[Full version]

Enumerating Reduced Polyominoes with Fixed Perimeter

A \emph{polyomino} is a shape best described as a connected set of cells in the square lattice. As part of recreational mathematics, polyominoes have seen active research since the 1950s. Simultaneously, polyominoes have been investigated in statistical physics under the name ``lattice animals,'' mainly in regards to percolation problems. One of the main points of interest is to solve the yet unanswered question of how many different polyominoes exist. Most of the focus, so far, ...

[Full version]

[Full version]

Keyword Search in Deduplicated Storage Systems

Deduplication is a widely implemented technique in storagesystems to reduce overall storage costs, or effectively increaselogical capacity, by replacing redundant chunks of data with references. As deduplication becomes a core component of storage systems, there is an opportunity to rethink storage functions to leverage the properties of deduplication, that thelogical size of a storage system may be many multiples ofthe physical data size. Specifically, we focus on the common task of performing a search for ...

[Full version]

[Full version]

Matter of Perspective Course - First Collaboration between Computer Science and Architecture

Monday, 12.7.2021, 10:00

CS Taub Lobby

You are invited to projects ipresentation in The Matter of Perspective course today, Monday, July 12, 2021, between 10: 00-11: 30 in the Taub lobby. This is a unique and first course shared by the Faculty of Computer Science and the Faculty of Architecture in which students work in mixed groups in order to produce a physical product in digital production technologies, using a geometric algorithm implementation. The course is conducted by Prof. Gershon Elber ...

[Full version]

[Full version]

Leveraging World Events to Predict E-Commerce Consumer Demand under Anomaly

Consumer demand forecasting is of high importance for many e-commerce applications, including supply chain optimization, advertisement placement, and delivery speed optimization. However, reliable time series sales forecasting for e-commerce is difficult, especially during periods with many anomalies, as can often happen during pandemics, abnormal weather, or sports events. Although many time-series algorithms have been applied to the task, prediction during anomalies still remains a challenge. In this work, we hypothesize that leveraging external knowledge found ...

[Full version]

[Full version]

Deep Generative Models for Molecular Optimization

Molecular lead optimization is an important task of drug discovery focusing on generating novel molecules similar to a drug candidate but with enhanced properties. Prior works focused on supervised models requiring datasets of pairs of a molecule and an enhanced molecule. These approaches require large amounts of data and are limited by the bias of the specific examples of enhanced molecules. In this Thesis, we first tackle the molecule optimization problem and present an unsupervised ...

[Full version]

[Full version]

Sequence Reconstruction Problem

Binary and q-ary sequences have always been used in communication channel as the carrier or the vessel of information. In order to establish an efficient and error-free communication channel, investigations on the properties of sequences are crucial. The property that we will investigate in this seminar is the reconstruction capability of binary sequences in particular from its subsequences. This is called the Sequence Reconstruction Problem. The problem considers a communication scenario where the sender transmits ...

[Full version]

[Full version]

A Generative Approach for Mitigating Structural Biases in Natural Language Inference

Many natural language inference (NLI) datasets contain biases that allow models to perform well by only using a biased subset of the input, without considering the remainder features. For instance, models are able to make a classification decision by only using the hypothesis, without learning the true relationship between it and the premise. These structural biases lead discriminative models to learn unintended superficial features and to generalize poorly out of the training distribution. In this ...

[Full version]

[Full version]

Project Fair in IoT, Software, Android Apps, AI, Cyber, Computer Security, and Networks

Tuesday, 29.6.2021, 12:30

CS Taub Lobby

CS Labs: Systems and Software Development Laboratory (SSDL), Cyber and Computer Security Laboratory (CYBER), The Laboratory for Computer Communication and Networking (LCCN) invite you to visit the Spring Project Fair in IoT, Software, Android Apps, AI, Cyber, Computer Security, and Networks, including demos and presentations by 40 undergraduate teams who will answer your questions on their research. The event will be held on Tuesday, June 29, 2021, at 12, in the CS Taub Lobby. You ...

[Full version]

[Full version]

CGGC Seminar: Hyperspectral Inverse Skinning

In example-based inverse linear blend skinning (LBS), a collection of poses (e.g., animation frames) are given, and the goal is finding skinning weights and transformation matrices that closely reproduce the input. These poses may come from physical simulation, direct mesh editing, motion capture, or another deformation rig. We describe a re-formulation of inverse skinning as a problem in high-dimensional Euclidean space. The transformation matrices applied to a vertex across all poses can be thought of ...

[Full version]

[Full version]

Distributed Computing Seminar: A Fully Adaptive Self-Stabilizing Transformer for LCL Problems

Shimon Biton (IE, Technion)

Sunday, 27.6.2021, 11:30

Zoom Lecture: 99794260392 and Bloomfield 152 (Hybrid manner)

The first generic self-stabilizing transformer for local problems in a constrained bandwidth model is introduced. This transformer can be applied to a wide class of locally checkable labeling (LCL) problems, converting a given fault free synchronous algorithm that satisfies certain conditions into a self-stabilizing synchronous algorithm for the same problem. The resulting self-stabilizing algorithms are anonymous, size-uniform, and \emph{fully adaptive} in the sense that their time complexity is bounded as a function of the number ...

[Full version]

[Full version]

ceClub: Operating Systems Abstractions for Trusted Execution Environments

Trusted execution environments such as secure enclaves are now available in several popular CPUs, and supported in public clouds. Enclaves can be used to efficiently shield applications against privileged adversaries, and secure sensitive data processed by them through strong isolation backed by the hardware. Yet, enclaves are not a silver bullet: they are vulnerable to unique side-channel attacks, they exhibit poor performance when system calls are invoked and when page faults occur, they lack a ...

[Full version]

[Full version]

Recruitment Day by Vayyar

Wednesday, 23.6.2021, 10:00

CS Taub Lobby

Vayyar representatives will visit CS to demonstrate their Radar-based technological solutions, on Wednesday, June 23, 2021, between 10:00-17:00, at the CS Taub Lobby. More details in the attached poster. You are all invited!

[Full version]

[Full version]

Domain Adaptation with Category Shift, an Application to Aspect Extraction

The rise of pre-trained language models has yielded substantial progress in the vast majority of Natural Language Processing (NLP) tasks. However, a generic approach towards the pre-training procedure can naturally be sub-optimal in some cases. Particularly, fine-tuning a pre-trained language model on a source domain and then applying it to a different target domain, results in a sharp performance decline of the eventual classifier for many source-target domain pairs. Moreover, in some NLP tasks, the ...

[Full version]

[Full version]

Pixel Club: The Trimmed Lasso: Sparse Recovery Guarantees and Practical Optimization

Consider the sparse approximation or best subset selection problem: Given a vector y and a matrix A, find a k-sparse vector x that minimizes the residual ||Ax-y||. This sparse linear regression problem, and related variants, plays a key role in high dimensional statistics, compressed sensing, machine learning and more. In this talk we focus on the trimmed lasso penalty, defined as the L_1 norm of x minus the L_1 norm of its top k entries ...

[Full version]

[Full version]

Analysis of Two-variable Recurrence Relations with Application to Parameterized Approximations

We introduce randomized branching as a tool for parameterized approximation and develop the mathematical machinery for its analysis. Our algorithms substantially improve the best known running times of parameterized approximation algorithms for Vertex Cover and $3$-Hitting Set for a wide range of approximation ratios. The running times of our algorithms are derived from an asymptotic analysis of a broad class of two-variable recurrence relations. Our main theorem gives a simple formula for this asymptotics. The ...

[Full version]

[Full version]

Meetup by RAFAEL

Wednesday, 16.6.2021, 18:00

Nola Socks Pub, Haifa

Rafael will hold a Meetup meeting with the participation of Gidi Weiss, VP of Marketing and Business Development in the division, who will talk about the most advanced security technologies in the world. The meeting will take place on Wednesday, May 26, 2021, at the Nola Socks Pub, Haifa, and participation requires pre-registration.

[Full version]

[Full version]

Best Project Contest - The Finals

Wednesday, 16.6.2021, 12:30

CS Taub Lobby

You are invited to the finals event of the Best Project Contest, to be held on Wednesday, June 16, 2021, starting at 12:30 and at 14:00 announcing and awarding the winners, at the CS Taub Lobby. The event will take place in the format of a project fair, and in accordance with the guidelines of the green pass instructions. You are all invited to come and meet the best researchers and researches!

[Full version]

[Full version]

DLACEP: A Deep-Learning Based Framework for Approximate Complex Event Processing

Complex event processing (CEP) is employed to detect user-specified patterns of events in data streams. CEP mechanisms operate by maintaining all sets of events that can potentially be composed into a pattern match. This approach can be wasteful when many of the sets do not participate in an actual match and are therefore discarded. We present DLACEP, a novel framework that fuses deep learning with CEP to efficiently extract complex pattern matches from streams. To ...

[Full version]

[Full version]

CGGC Seminar: Discrete Willmore Surfaces

Olga Diamanti (TU Graz, Institute for Geometry)

Wednesday, 16.6.2021, 11:30

Zoom Lecture: 91344952941

This talk will be about the problem of discrete constrained Willmore surfaces: discrete surfaces that have minimal total squared mean curvature while also being discretely conformally equivalent to a given input surface. The Willmore energy is a bending energy, used to model elastic behavior and measure surface smoothness. Adding the conformality constraint turns the problem into a natural extension, in 2D, of classical elastic spline modeling in 1D. This not only makes the use of ...

[Full version]

[Full version]

Numerical Optimization and Multigrid Computational Methods with Applications

Work 1: we introduce a way to adapt Nesterov's well-known scheme to accelerating stationary iterative solvers for linear systems. Compared with classical Krylov subspace acceleration methods, the proposed scheme requires more iterations, but it is trivial to implement and retains essentially the same computational cost as the unaccelerated method. An explicit formula for a fixed optimal parameter is derived in the case where the stationary iteration matrix has only real eigenvalues, based only on the ...

[Full version]

[Full version]

Pixel Club: Subsampled Brain MRI Reconstructionby Generative Adversarial Neural Networks

A main challenge in magnetic resonance imaging (MRI) is speeding up scan time. Beyond improving patient experience and reducing operational costs, faster scans are essential for time-sensitive imaging, such as fetal, cardiac, or functional MRI, where temporal resolution is important and target movement is unavoidable, yet must be reduced. Current MRI acquisition methods speed up scan time at the expense of lower spatial resolution and costlier hardware. We introduce a practical, software-only framework, based on ...

[Full version]

[Full version]

Scaling Up Board Games with AlphaZero and Graph Neural Networks

Playing board games is considered a major challenge for both humans and AI researchers. Because some complicated board games are quite hard to learn, humans usually begin with playing on smaller boards and incrementally advance to master larger board strategies. Most neural network frameworks that are currently tasked with playing board games neither perform such incremental learning nor possess capabilities to automatically scale up. In this work, we look at the board as a graph ...

[Full version]

[Full version]

Pixel Club: Scene Understanding by Iterative Bottom-up top-down Processing

Prof. Shimon Ullman (Weizmann Institute of Science)

Tuesday, 8.6.2021, 11:30

Zoom Lecture: 99725686717

Scene understanding requires the extraction and representation of scene components together with their individual properties, as well relations and interactions between them. In current computer vision, there has been considerable progress in recognizing scene components (people, objects, parts), but the problem of recovering scene structure is still largely open. I will describe a model that performs scene interpretation by an iterative process, combining bottom-up and top-down networks, interacting through a symmetric bi-directional communication between them. ...

[Full version]

[Full version]

Machine Learning for Programming Language Processing

This talk will focus on structural representations and neural models of source code. I will present a language-agnostic approach for structural language modeling (SLM) of code. This general approach obtains state-of-the-art results in a variety of tasks including code summarization, code captioning, code completion, name prediction, and edit completion, outperforming sequence models (such as textual Transformers and LSTMs) and models based on graph neural networks (GNNs). Studying the reason why GNNs do poorly compared to ...

[Full version]

[Full version]

Online Meeting with Microsoft

Monday, 7.6.2021, 18:30

TEAMS Event: Registration

You are invited to an online meeting (TEAMS) with Microsoft representatives and to hear from their students on the work experience in the company and the combination of studies and careers, from the managers and the recruitment team on job interviews, and more, on Monday, June 7, 202, 18:30. A link to the meeting will be sent upon pre-registration.

[Full version]

[Full version]

CGGC Seminar: Neural 3D Reconstruction

Neural networks have made exciting progress on unstructured 3D geometric data; which is changing the way we fundamentally approach problems in geometry processing. In this talk, I will discuss several works which facilitate 3D reconstruction from several different directions, including consolidating point clouds, estimating a globally consistent point normal orientation, and reconstructing a surface mesh. Finally, I will conclude with ongoing and future work in this direction, as well as other related areas. The lecture ...

[Full version]

[Full version]

Online Lecture on the Way from Taub to Google Japan

Thursday, 3.6.2021, 16:30

Zoom Event: Registration

You are invited to an online lecture by Sarai Duak, a CS graduate and currently a Data Scientist Lead at Google Tokyo, Japan, on simple solutions for business development problems for customers, with the help of Data Science, on Thursday, June 3, 2021, 16:30. Link to the Zoom meeting will be sent upon pre-registration. More details in the attached poster.

[Full version]

[Full version]

Pixel Club: A New Theory of Adversarial Examples in Machine Learning

The extreme fragility of deep neural networks when presented with tiny perturbations in their inputs was independently discovered by several research groups in 2013. Due to their mysterious properties and major security implications, these adversarial examples had been studied extensively over the last eight years, but in spite of enormous effort they remained a baffling phenomenon with no clear explanation. In particular, it was not clear why a tiny distance away from almost any cat ...

[Full version]

[Full version]

ML Based Lineage in Databases

There has been extensive research on data provenance. Previous works were concerned with annotating the results of database (DB) queries with provenance information which is useful in explaining query results at various resolution levels. In this work, we track the lineage of tuples throughout their database lifetime. That is, we consider a scenario in which tuples (records) that are produced by a query may affect other tuple insertions into the DB, as part of a ...

[Full version]

[Full version]

Online Meeting with Intel on Job Interviews

Monday, 31.5.2021, 18:00

Zoom Event: Registration

You are invited to an online meeting with Intel representatives and to hear from their software engineers on the work experience as interviewers, including tools, exercising and tips for success in the technical stage of an interview, on Monday, May 31, 2021, 18:00. Link to the meeting will be sent upon pre-registration.

[Full version]

[Full version]

Neural Algorithms for Precise Shape Completion

According to Aristotle, “the whole is greater than the sum of its parts”. This statement was adopted to explain human perception by the Gestalt psychology school of thought in the twentieth century. Here, we claim that when observing a part of an object which was previously acquired as a whole, one could deal with both partial correspondence and shape completion in a holistic manner. More speciﬁcally, given the geometry of a full, articulated object in ...

[Full version]

[Full version]

Unintended Features of APIs: Cryptanalysis of Incremental HMAC

Many cryptographic APIs provide extra functionality that was not intended by the designers. In this seminar we discuss such an unintended functionality in the API of HMAC as implemented by Siemens and OpenSSL. HMAC authenticates a single message at a time with a single authentication tag. However, most HMAC implementations do not complain when extra data is added to the stream after a tag is computed. We call such primitives Incremental MACs. Though HMAC is ...

[Full version]

[Full version]

ceClub: Machine Learning in Compiler Optimization

The end of Moore's law is driving the search for new techniques to improve system performance as applications continue to evolve rapidly and computing power demands continue to rise. One promising technique is to build more intelligent compilers. Compilers map high-level programs to lower-level primitives that run on hardware. During this process, compilers perform many complex optimizations to boost the performance of the generated code. These optimizations often require solving NP-Hard problems and dealing with ...

[Full version]

[Full version]

Security of Quantum Key Distribution Protocols

Rotem Liss

Wednesday, 5.5.2021, 11:30

Zoom Lecture:
99607663751

For password to lecture, please contact: rotemliss@cs.technion.ac.il

For password to lecture, please contact: rotemliss@cs.technion.ac.il

The counter-intuitive features of quantum mechanics make it possible to solve problems and perform tasks that are beyond the abilities of non-quantum (classical) computers and communication devices. In particular, quantum key distribution (QKD) protocols allow two participants (Alice and Bob) to achieve the classically-impossible task of generating a secret shared key even if their adversary is computationally unlimited. Unfortunately, the security promises of QKD are true only in theory; practical implementations of QKD deviate from ...

[Full version]

[Full version]

ceClub: The Technion Computer Engineering Club

In Modern data centers, resources are usually virtualized. Applications running on those date centers are distributed over many virtual machines. For those applications, the data centers provide software defined infrastructure services for networking, storage, and security. When software defined services are running within the same CPU as the applications, they consume CPU resources on the expanse of the applications. Moreover, the data center security can be jeopardized NVIDIA Data Center Infrastructure Processing Unit (DPU) allow ...

[Full version]

[Full version]

Intel Ergonomics Workshop to Upgrade the Study Position

Monday, 3.5.2021, 17:00

Zoom Event: Registration

Intel representative, an expert in ergonomics, will hold a workshop on the subject on Monday, May 3, 2021, at 17:00, with tips for upgrading the distance learning environment. A link to Participants will be sent after pre-registration.

[Full version]

[Full version]

Geometrical Challenges in Treating Irregular Heart Beat

Fady Massarwi (CS, Technion)

Monday, 3.5.2021, 11:30

Zoom Lecture: 91344952941
For password to lecture please contact inbalb@cs.technion.ac.il

This talk presents some of the geometrical aspects involved in treating irregular heart beat rhythm (Arrythmia) using Carto 3 System. Carto 3 is a product of Biosense-Webster, a global leader in the science of diagnosing and treating heart rhythm disorders. CARTO 3 System enables accurate visualization of multiple catheters in a patient’s heart and pinpoints exact location/orientation of a catheter. During arrythmia procedure, a 3D electro-anatomical reconstruction of the heart is built and color coded ...

[Full version]

[Full version]

Lecture on Quantum Calculation: What is it and why is it Cool?

Thursday, 29.4.2021, 17:00

Zoom Event: Registration

You are invited to a lecture on quantum computing: what is it and why is it cool? by Dr. Gadi Alexandrovich - CS graduate, a researcher in the IBM research laboratory in Haifa in the field of quantum computing and the author of the "Inaccurate" mathematical blog - which will deal with quantum computers and the changes they will bring about in the future. The lecture will take place on Thursday, April 29, 17:00, in ...

[Full version]

[Full version]

Computational inference of cancer metabolic alterations for early diagnosis and treatment

Shoval Lagziel

Thursday, 29.4.2021, 15:30

Zoom Lecture:
93506187830

For password to lecture, please contact: shovall@cs.technion.ac.il

For password to lecture, please contact: shovall@cs.technion.ac.il

Metabolic reprogramming is a hallmark of cancer, providing novel means to selectively target cancer cells, for precision medicine and early diagnosis. Understanding tumor-specific metabolic alterations facilitates the identification of induced dependency on specific enzymes whose inhibition selectively targets cancer cells. In addition, the altered metabolic activity of cancer cells, involving the consumption of metabolic nutrients and the secretion of byproducts from the tumor leaves metabolic traces that can be utilized for diagnostic purposes. Here, we ...

[Full version]

[Full version]

Maximizing Throughput in Flow Shop Real-time Scheduling

Lior Ben-Yamin

Thursday, 29.4.2021, 14:30

Zoom Lecture:
93508538152

For password to lecture, please contact: lior.b@cs.technion.ac.il

For password to lecture, please contact: lior.b@cs.technion.ac.il

We consider scheduling real-time jobs in the classic flow shop model. The input is a set of n jobs, each consisting of m segments to be processed on m machines in the specified order. Each job also has a release time, a due date, and a weight. The objective is to maximize the throughput, i.e., to find a subset of the jobs that have the maximum total weight and can complete processing on the m ...

[Full version]

[Full version]

Reshaping the Roles of Humans and Al in Data Integration

Roee Shraga - Guest Lecture

Tuesday, 27.4.2021, 12:30

HYBRID - Taub 5 (Green Pass) and
Zoom Lecture:
91488539030

The matching task is at the heart of data integration, in charge of aligning elements of data sources. Matching is a handy tool in multiple contemporary business and commerce applications and has been investigated in the fields of databases, AI, Semantic Web, and data mining for many years. The core challenge still remains the ability to create quality algorithmic matchers, automatic tools for identifying correspondences among data concepts (e.g., database attributes). Matching problems were traditionally ...

[Full version]

[Full version]

Pixel Club: On the Connection between Deep Neural Networks and Kernel Methods

Recent theoretical work has shown that massively overparameterized neural networks are equivalent to kernel regressors that use Neural Tangent Kernels (NTKs). Experiments indicate that these kernel methods perform similarly to real neural networks. My work in this subject aims to better understand the properties of NTK and relate them to properties of real neural networks. In particular, I will argue that for input data distributed uniformly on the sphere NTK favors low frequency predictions over ...

[Full version]

[Full version]

git Workshop at CS

Wednesday, 21.4.2021, 17:30

Zoom Event: Registration

You are invited to an online technological workshop by Aviv Rosenberg,CS Ph.D. student and TA, on versioning with git: How to stop being afraid of changing code, on Wednesday, April 21, 2021, 17:30. More details on the the workshop agenda and pre-registration.

[Full version]

[Full version]

PCPs and Cryptography: New Limitations and Opportunities

Liron Bronfman

Wednesday, 21.4.2021, 14:00

Zoom Lecture:
5480679598

For password to lecture, please contact: br@cs.technion.ac.il

For password to lecture, please contact: br@cs.technion.ac.il

The connection between information theoretic proof systems and cryptography has been extremely fruitful. In this thesis, we further explore this connection, showing both new limitations and opportunities. In the talk we will focus on the new opportunities and show constructions of computational relaxations of objects that are known to be essentially impossible to achieve information theoretically. In particular, we show cryptographic analogs of: (1) PCPs whose length is proportional to the witness size. (2) Instance ...

[Full version]

[Full version]

CS Open Day for Graduate Studies

Wednesday, 21.4.2021, 12:30

Zoom Event: Registration

Technion CS open day 2021 invites outstanding undergraduates from all universities to learn about the Computer Science Department and register for Winter Semester 2021-22. The event will be held online by ZOOM - ID MEETING NO. 96244586510, on Wednesday, April 21, 2021. between 12:30-13:45. The program will include review on curriculum, research and life at the Technion CS Department: 12:30-12:40 CS Dean, Prof. Dan Geiger 12:40-12:55 Vice Dean, Prof. Gill Barequet 12:55-13:20 Dr. Kira Radinsky, ...

[Full version]

[Full version]

Recruitment Day By CISCO

Monday, 19.4.2021, 18:00

Zoom Event: Registration

CISCO will hold an online recruitment on Monday, April 19th, 2021, 18:00, including meetings with the company's students and engineers who will tell you about working at Cisco's various locations and about openings. More details, and registration. You are all invited!

[Full version]

[Full version]

CGGC Seminar: Topological and Geometric Analysis of Graphs

Yusu Wang (University of California)

Monday, 19.4.2021, 17:00

Zoom Lecture: 91344952941
For password to lecture please contact inbalb@cs.technion.ac.il

In recent years, topological and geometric data analysis (TGDA) has emerged as a new and promising field for processing, analyzing and understanding complex data. Indeed, geometry and topology form natural platforms for data analysis, with geometry describing the ”shape” behind data; and topology characterizing / summarizing both the domain where data are sampled from, as well as functions and maps associated to them. In this talk, I will show how topological (and geometric ideas) can ...

[Full version]

[Full version]

Data Science & Deep Learning: State Visitation Fairness in Average-Reward MDPs

Vineet Nair (CS, Technion)

Monday, 19.4.2021, 12:30

Zoom Lecture:
93378688224

For password to lecture, please contact: mayasidis@cs.technion.ac.il

For password to lecture, please contact: mayasidis@cs.technion.ac.il

Fairness has emerged as an important concern in automated decision-making in recent years, especially when these decisions affect human welfare. In this work, we study fairness in temporally extended decision-making settings, specifically those formulated as Markov Decision Processes (MDPs). Our proposed notion of fairness ensures that each state's long-term visitation frequency is more than a specified fraction. In an average-reward MDP setting, we formulate the problem as a bilinear saddle point program and, for a ...

[Full version]

[Full version]

Deep Generative Models for ECG Classification

Tomer Golany

Sunday, 18.4.2021, 09:00

Zoom Lecture:
996761764160

For password to lecture, please contact: tomer.golany@cs.technion.ac.il

For password to lecture, please contact: tomer.golany@cs.technion.ac.il

32% of all global deaths in the world are caused by cardiovascular diseases. The Electrocardiogram (ECG) is a non-invasive tool to measure the electrical activity of the heart, and it is the most common test performed by cardiologists to detect heart-diseases. Analyzing ECG signals manually is a hard task. Furthermore, abnormalities in the heart may occur at any time and not necessarily in the hospital. Many attempts were made to automate this task using machine ...

[Full version]

[Full version]

Sparse Linear Networks with a Fixed Butterfly

Omer Leibovitch

Monday, 12.4.2021, 12:30

Zoom Lecture:
98712430421

For password to lecture, please contact: mayasidis@campus.technion.ac.il

For password to lecture, please contact: mayasidis@campus.technion.ac.il

A butterfly network consists of logarithmically many layers, each with a linear number of pre-specified nonzero weights. We propose to replace a dense linear layer in any neural network by an architecture based on the butterfly network. The proposed architecture significantly improves upon the quadratic number of weights required in a standard dense layer to nearly linear with little compromise in expressibility of the resulting operator. In a collection of wide variety of experiments, including ...

[Full version]

[Full version]

Data Science & Deep Learning: Sparse Linear Networks with a Fixed Butterfly

Omer Leibovitch (CS, Technion)

Monday, 12.4.2021, 12:30

Zoom Lecture:
98712430421

For password to lecture, please contact: mayasidis@cs.technion.ac.il

For password to lecture, please contact: mayasidis@cs.technion.ac.il

A butterfly network consists of logarithmically many layers, each with a linear number of pre-specified nonzero weights. We propose to replace a dense linear layer in any neural network by an architecture based on the butterfly network. The proposed architecture significantly improves upon the quadratic number of weights required in a standard dense layer to nearly linear with little compromise in expressibility of the resulting operator. In a collection of wide variety of experiments, including ...

[Full version]

[Full version]

CGGC Seminar: Deep 3D Generative Modeling

Niloy J. Mitra (University College London (UCL))

Monday, 12.4.2021, 10:30

Zoom Lecture: 91344952941
For password to lecture please contact inbalb@cs.technion.ac.il

Deep learning has taken the Computer Graphics world by storm. While remarkable progress has been reported in the context of supervised learning, the state of unsupervised learning, in contrast, remains quite primitive. In this talk, we will discuss recent advances where we have combined knowledge from traditional computer graphics and image formation models to enable deep generative modeling workflows. We will describe how we have combined modeling and rendering, in the unsupervised setting, to enable ...

[Full version]

[Full version]

Code Retreat Workshop at CS

Sunday, 11.4.2021, 17:30

Zoom Event: Registration

You are invited to participate in the Code Retreat workshop that will take place at CS for the first time, in order sharpen the development skill and practice a four-hand programming method in four hands and one keyboard (Pair Programming), during which participants practice writing code in pairs and sharpen code skills while coordinating group work expectations, dealing with one thought, explore and practice different methods of software development: • Starting with a simple programming ...

[Full version]

[Full version]

Complex Event Forecasting in Multivariate Time Series

Dolev Elbaz

Sunday, 11.4.2021, 11:00

Zoom Lecture:
996692671429

For password to lecture, please contact: dolevelb@campus.technion.ac.il

For password to lecture, please contact: dolevelb@campus.technion.ac.il

Time-series forecasting is widely employed in a variety of domains to predict future trends, tendencies, and properties of the data. However, predicting simple data items is often not enough. Many applications are characterized by a requirement to simultaneously monitor hundreds or even thousands of data series and could benefit from recognizing future occurrences of composite patterns in advance. Despite the rising need for such functionality, this problem received limited attention in recent years. In this ...

[Full version]

[Full version]

What If: Answer Simulation Questions by Generating Code

Gal Peretz

Thursday, 8.4.2021, 16:30

Zoom Lecture:
98204535821

For password to lecture, please contact: sgalprz@cs.technion.ac.il

For password to lecture, please contact: sgalprz@cs.technion.ac.il

Many texts, especially in Chemistry and Biol-ogy, describe complex processes. To answer questions about such processes one needs to understand the interactions between the different entities and to track the state transition between the different stages of the process. In this work, we tackle this problem by learning to generate corresponding code to a text that describes a chemical reaction process and a question that asks about the process outcome in a different setup. We ...

[Full version]

[Full version]

On Anomaly Detection in Tabular Data

Igor Margulis

Wednesday, 7.4.2021, 11:30

Zoom Lecture:
91383403107

For password to lecture, please contact: margulis@campus.technion.ac.il

For password to lecture, please contact: margulis@campus.technion.ac.il

Anomaly detection is a technique for finding unusual patterns in the given data. The study of anomaly detection has a long history and spans multiple disciplines including engineering, machine learning, statistics and real-life applications. We consider the problem of anomaly detection in tabular data, and present a detection scheme which is based on training a multiway classification model for discriminating between dozens of transformations applied to given "normal" records. The auxiliary expertise learned by the ...

[Full version]

[Full version]

A Structural Model for Contextual Code Changes

Shaked Brody

Tuesday, 6.4.2021, 11:00

Zoom Lecture:
96914709680

For password to lecture, please contact: shakedbr@cs.technion.ac.il

For password to lecture, please contact: shakedbr@cs.technion.ac.il

We address the problem of predicting edit completions based on a learned model that was trained on past edits. Given a code snippet that is partially edited, our goal is to predict a completion of the edit for the rest of the snippet. We refer to this task as the Edit Completion task and present a novel approach for tackling it. The main idea is to directly represent structural edits. This allows us to model ...

[Full version]

[Full version]

Pixel Club: Deep Networks from First Principles

In this talk, we offer an entirely “white box’’ interpretation of deep (convolution) networks from the perspective of data compression (and group invariance). In particular, we show how modern deep layered architectures, linear (convolution) operators and nonlinear activations, and even all parameters can be derived from the principle of maximizing rate reduction (with group invariance). All layers, operators, and parameters of the network are explicitly constructed via forward propagation, instead of learned via back propagation. ...

[Full version]

[Full version]

The Deletion/Insertion Channel and its Application to Coding for DNA Storage

Daniella Bar-Lev

Monday, 5.4.2021, 17:30

Zoom Lecture:
97428473352

For password to lecture, please contact: daniellalev@cs.technion.ac.il

For password to lecture, please contact: daniellalev@cs.technion.ac.il

DNA-based storage offers significant advantages over magnetic and optical storage solutions in terms of density, durability and not requiring a constant power supply. Given current trends of cost reduction in DNA synthesis and sequencing, it is now acknowledged that within the next 10 – 15 years DNA-based storage may become a highly competitive archiving technology. The microscopic world in which the DNA molecules reside induces error patterns that are fundamentally different from their digital counterparts. ...

[Full version]

[Full version]

Recruitment Day By NVDIA

Monday, 5.4.2021, 17:30

TEAMS Event

NVDIA will hold a TEAMS online recruitment day today, Monday, April 5th, 2021, including short technological lectures and Q&A meeting with the company's engineers, job offers and information about openings. You are all invited.

[Full version]

[Full version]

Data Science & Deep Learning: Coresets for Some Machine Learning Algorithms

Supratim Shit (Indian Institute of Technology Gandhinagar)

Monday, 5.4.2021, 12:30

Zoom Lecture:
98095992835

A butterfly network consists of logarithmically many layers, each with a linear number of pre-specified nonzero weights. We propose to replace a dense linear layer in any neural network by an architecture based on the butterfly network. The proposed architecture significantly improves upon the quadratic number of weights required in a standard dense layer to nearly linear with little compromise in expressibility of the resulting operator. In a collection of wide variety of experiments, including ...

[Full version]

[Full version]

Leveraging Drug Modalities for Drug Repurposing

Galia Nordon

Sunday, 4.4.2021, 11:00

Zoom Lecture:
96920869630

For password to lecture, please contact: galiasn@cs.technion.ac.il

For password to lecture, please contact: galiasn@cs.technion.ac.il

Drug repurposing is the process of applying known drugs to treat new diseases. Successful repurposing can reduce costs and time to market as medications have already passed studies of human safety. It is an important task due to the length of time and the large cost of novel drug development. In recent years, alongside the growing resources needed for developing new drugs, large biomedical repositories are becoming available as well as the maturing technology for ...

[Full version]

[Full version]

Reconstruction of Strings from their Substrings Spectrum

Sagi Marcovich

Wednesday, 24.3.2021, 16:00

Zoom Lecture:
92990701982

For password to lecture, please contact: sagimar@cs.technion.ac.il

For password to lecture, please contact: sagimar@cs.technion.ac.il

Using DNA molecules as a data storage volume was first introduced in the 1960s by Richard Feynman. Later, in 1990, the human genome project led to a significant progress in sequencing and assembly methods. As a result, the interest in storage solutions based on DNA molecules was increased. DNA storage enjoys major advantages over magnetic and optical storage solutions. Motivated by rising technologies for DNA sequencing, this work studies reconstruction of strings based upon their ...

[Full version]

[Full version]

CGGC Seminar: Geometric Construction of Auxetic Metamaterials

Stefanie Hahmann (University Grenoble INP)

Monday, 22.3.2021, 11:00

Zoom Lecture: 91344952941
For password to lecture please contact inbalb@cs.technion.ac.il

Recent advances in digital manufacturing, where computational design, materials science and engineering meet, offer whole new perspectives for tailoring mechanical properties and fabrication of material with applications as diverse as product design, architecture, engineering and art. Auxetic materials are characterized by a negative Poisson’s ratio. This means that they do not behave as usual materials. When stretched in one direction, they do not shrink in the other directions, in contrary they expand. In comparison to ...

[Full version]

[Full version]

A Meeting on Webinar by Huawei

Sunday, 21.3.2021, 19:30

Zoom Event: Registration

CS graduate studies students are invited to a meeting on Webinar by Huawei, on Sunday, March 21, 2021, 19:30-20:30/ For participation please pre-register by email. More details

[Full version]

[Full version]

To Foresee the Future - The Prediction that will Save the World - A Lecture by Dr. Kira Radinsky

Sunday, 21.3.2021, 17:00

Zoom Event: Registration

You are invited to a lecture by Dr. Kira Radinsky: "To Foresee the Future - The Prediction that will Save the World", on Sunday, March 21, 17:00. A link to the Zoom meeting will be sent upon pre-registration.

[Full version]

[Full version]

The Shapley Value of Tuples in Query Answering

Moshe Sebag

Sunday, 21.3.2021, 11:00

Zoom Lecture:
8029792183

For password to lecture, please contact: moshesebag@cs.technion.ac.il

For password to lecture, please contact: moshesebag@cs.technion.ac.il

This research aims to investigate the application of the Shapley value to quantify the contribution of a tuple to a query answer. The Shapley value is a widely known numerical measure in cooperative game theory and in many applications of game theory for assessing the contribution of a player to a coalition game. It has been established already in the 1950s, and is theoretically justified by being the very single wealth distribution measure that satisfies ...

[Full version]

[Full version]

Complex Pattern Mining

Eitan Kosman

Wednesday, 17.3.2021, 14:30

Zoom Lecture:
96049971966

For password to lecture, please contact: eitan.k@cs.technion.ac.il

For password to lecture, please contact: eitan.k@cs.technion.ac.il

Mining complex patterns from large data sets has attracted much attention in the last few decades. A plethora of methods and algorithms have been designed for mining a variety of patterns, ranging from simple association rules and frequent itemsets to advanced graph-based structures. However, as modern applications grow dramatically more sophisticated and operate on highly multidimensional and increasingly complex data, they introduce the demand for mining even more expressive and convoluted patterns unsupported by the ...

[Full version]

[Full version]

Shape correspondence by aligning scale-invariant LBO eigenfunctions

Amit Bracha

Tuesday, 16.3.2021, 11:30

Zoom Lecture:
3615145651

For password to lecture, please contact: amitbracha@cs.technion.ac.il

For password to lecture, please contact: amitbracha@cs.technion.ac.il

When matching non-rigid shapes, the regular or scale-invariant Laplace-Beltrami Operator (LBO) eigenfunctions could potentially serve as intrinsic descriptors which are invariant to isometric transformations. However, the computed eigenfunctions of two quasi-isometric surfaces could be substantially different. Such discrepancies include sign ambiguities and possible rotations and reflections within subspaces spanned by eigenfunctions that correspond to similar eigenvalues. Thus, without aligning the corresponding eigenspaces it is difficult to use the eigenfunctions as descriptors. In this talk, we ...

[Full version]

[Full version]

Investigating the Difference Between Emulated and Paravirtual Network I/O: The Strange, Untold Story

Aviv Ben-David

Monday, 15.3.2021, 18:00

Zoom Lecture:
96844553386

For password to lecture, please contact: bdaviv@cs.technion.ac.il

For password to lecture, please contact: bdaviv@cs.technion.ac.il

In virtual setups, guest virtual machines (VMs) perform their I/O through virtual I/O devices that are implemented by the hypervisor in software. There are two major flavors of virtual I/O devices. The first is ``emulation’’, which provides an interface identical to that of some preexisting physical I/O device, thus allowing the operating system (OS) inside the VM to use the original driver of the device, as is, unaware that it is in fact virtual (implemented ...

[Full version]

[Full version]

Extracting Bible Quotes from Historical Commentary

Asaf Yeshurun

Sunday, 14.3.2021, 11:00

Zoom Lecture:
5201760342

For password to lecture, please contact: asafyeshurun@cs.technion.ac.il

For password to lecture, please contact: asafyeshurun@cs.technion.ac.il

The Hebrew Bible (Tanach) has been extensively quoted by historical religious text and commentaries throughout history. Nowadays, many of these text resources are publicly available online. Yet, the Bible quotations within them are often partially identified if at all. Knowing the exact quotations may be highly beneficial to scholars interested in studying or investigating the Bible. We have developed and empirically analyzed a machine-learning solution for this task. End-to-end, our model is comprised of three ...

[Full version]

[Full version]

Cognitive Models in Deep Learning

Idan Schwartz

Wednesday, 10.3.2021, 16:30

Zoom Lecture:
9855273458

For password to lecture, please contact: idansc@cs.technion.ac.il

For password to lecture, please contact: idansc@cs.technion.ac.il

The quest for algorithms that enable cognitive abilities is an integral part of machine learning and appears in many facets, such as virtual assistant and visual reasoning. A cognitive system requires an effective approach to extract details and nuances from the multiple sensors that pound the devices' computational engine. To this end, we propose a novel form of attention mechanism, namely Factor Graph Attention, that operates on any data utilities and differentiates useful signals from ...

[Full version]

[Full version]

ceClub: Dragonblood: Analyzing the Dragonfly Handshake of WPA3 and EAP-pwd

The WPA3 certification aims to secure home networks, while EAP-pwd is used by certain enterprise WiFi networks to authenticate users. Both use the Dragonfly handshake to provide forward secrecy and resistance to dictionary attacks. In this paper, we systematically evaluate Dragonfly's security. First, we audit implementations, and present timing leaks and authentication bypasses in EAP-pwd and WPA3 daemons. We then study Dragonfly's design and discuss downgrade and denial-of-service attacks. Our next and main results are ...

[Full version]

[Full version]

Verizon Media Internship Meetup

Tuesday, 9.3.2021, 17:00

Zoom Event: Registration

CS graduate studies students are invited to an Internship Meetup by Verizon Media, on Tuesday, Mach 9, 2021, 17:00. For participation please pre-register by email. More details

[Full version]

[Full version]

Clustering in the Network Data Plane

Or Goaz

Tuesday, 9.3.2021, 14:00

Zoom Lecture:
99911513639

For password to lecture, please contact: orgoaz@cs.technion.ac.il

For password to lecture, please contact: orgoaz@cs.technion.ac.il

Clustering is a basic machine learning task. In this task, a stream of input items needs to be grouped into clusters, such that all items classified into the same cluster are closer to each other than to items classified to other clusters. Each cluster is centered around a centroid point, which may either be given as a parameter, or must be learned during the process in the case of unsupervised online learning. This work studies ...

[Full version]

[Full version]

Pixel Club: Geometric Deep Learning: the Erlangen Programme of ML

Michael Bronstein (Imperial College London)

Tuesday, 9.3.2021, 11:30

Zoom Lecture: https://technion.zoom.us/j/94556114100

For nearly two millennia, the word "geometry" was synonymous with Euclidean geometry, as no other types of geometry existed. Euclid's monopoly came to an end in the 19th century, where multiple examples of non-Euclidean geometries were shown. However, these studies quickly diverged into disparate fields, with mathematicians debating the relations between different geometries and what defines one. A way out of this pickle was shown by Felix Klein in his Erlangen Programme, which proposed approaching ...

[Full version]

[Full version]

Small Circuits Imply Efficient Arthur-Merlin Protocols

Michael Ezra

Tuesday, 9.3.2021, 11:00

Zoom Lecture:
5617822865

For password to lecture, please contact: michaelezra@cs.technion.ac.il

For password to lecture, please contact: michaelezra@cs.technion.ac.il

We show a new connection between circuit lower bounds and interactive proofs in restricted computational models. Specifically, we focus on the frontier problem of whether a DNF augmented with an additional layer of parity (XOR) gates, can approximate the inner product function. We show that the existence of such a small circuit, would have unexpected general implications for interactive variants of the Data Streaming and Communication Complexity models.

[Full version]

[Full version]

Adversarial Examples for Models of Code and Defending Against Them

Noam Yefet

Thursday, 4.3.2021, 12:30

Zoom Lecture:
96898381897

For password to lecture, please contact: snyefet@cs.technion.ac.il

For password to lecture, please contact: snyefet@cs.technion.ac.il

Neural models of code have shown impressive results when performing tasks such as predicting method names and identifying certain kinds of bugs. We show that these models are vulnerable to adversarial examples, and introduce a novel approach for attacking trained models of code using adversarial examples. The main idea of our approach is to force a given trained model to make an incorrect prediction, as specified by the adversary, by introducing small perturbations that do ...

[Full version]

[Full version]

Automata over Infinite Data Domains: Learnability and Applications in Program Verification and Repair

Hadar Frenkel

Tuesday, 2.3.2021, 17:00

Zoom Lecture:
97090529670

For password to lecture, please contact: hfrenkel@cs.technion.ac.il

For password to lecture, please contact: hfrenkel@cs.technion.ac.il

We present automata over infinite data domains and their use in program verification and repair. In particular, we discuss assume-guarantee based verification, a compositional verification method that uses automata learning in order to modularly verify the correctness of a system. Then we present Assume-Guarantee-Repair (AGR) – a framework that verifies that a program satisfies a set of properties, and repairs the program in case the verification fails. We consider communicating programs – these are simple ...

[Full version]

[Full version]

Designing Deep Neural Networks for Efficient and Robust Inference

Chaim Baskin

Tuesday, 2.3.2021, 11:30

Zoom Lecture:
99572398109

For password to lecture, please contact: chaimbaskin@cs.technion.ac.il

For password to lecture, please contact: chaimbaskin@cs.technion.ac.il

Deep neural networks (DNN) became a common tool for solving complex tasks in various fields such as computer vision, natural language processing, and recommendation systems. Despite recent progress made in enhancing the DNN performance, there are still two major obstacles hindering the practicality of DNNs in some application: their energy-expensive deployment on embedded platforms, and their amenability to malicious adversarial perturbations. In this talk, I will overview several lines of works tackling different aspects of ...

[Full version]

[Full version]

Google Hash Code 2021

Thursday, 25.2.2021, 19:30

Zoom Event: Registration

Google Hash Code 2021will take place on Thursday, February 25, 2021 between 19:30-23:45 and you are invited to register to the Technion Hub by Wednesday, February 24, 13:00 IST. More details and registration. Technion Hub Facebook Group

[Full version]

[Full version]

Fault Tolerant Max-Cut

Noa Marelly

Thursday, 25.2.2021, 14:30

Zoom Lecture:
98844121807

For password to lecture, please contact: noa.marelly@cs.technion.ac.il

For password to lecture, please contact: noa.marelly@cs.technion.ac.il

In this work, we initiate the study of fault tolerant Max-Cut, where given an edge-weighted undirected graph G=(V,E), the goal is to find a cut S, that maximizes the total weight of edges that cross S even after an adversary removes k vertices from G. We consider two types of adversaries: an adaptive adversary that sees the outcome of the random coin tosses used by the algorithm, and an oblivious adversary that does not. For ...

[Full version]

[Full version]

Scalable deep learning with pipeline model parallelism

Saar Eliad

Thursday, 25.2.2021, 11:00

Zoom Lecture:
94960294313

For password to lecture, please contact: saareliad@cs.technion.ac.il

For password to lecture, please contact: saareliad@cs.technion.ac.il

We worked on a particular case of Deep Learning where the model is too large to fit into the memory of a single commodity GPU during training. Such is the case for fine-tuning, an increasingly common technique that leverages transfer learning to dramatically expedite the training of huge, high-quality models. Critically, it holds the potential to make giant state-of-the-art models pre-trained on high-end super-computing-grade systems readily available for users that lack access to such costly ...

[Full version]

[Full version]

Pixel Club: Learning like Humans Do, with Limited Training Data

Amit Alfassy (EE, Technion)

Tuesday, 23.2.2021, 11:30

Zoom Lecture: https://technion.zoom.us/j/95741652165

While Deep learning has brought a huge advancement to computer vision, for most tasks we still need hundreds of labeled samples per class. The few-shot learning tasks attempts to alleviate the data problem by learning from 1/ 5 samples per class. We will discuss the few-shot learning domain through two of my papers. The first paper LaSO, is a SOTA augmentation mechanic for multi-label few-shot classification and was published in CVPR 2019. The second paper ...

[Full version]

[Full version]

Batched Vertex Cover Reconfiguration

Shahar Romem Peled

Thursday, 18.2.2021, 14:30

Zoom Lecture:
99681314877

For password to lecture, please contact: shaharr@cs.technion.ac.il

For password to lecture, please contact: shaharr@cs.technion.ac.il

Our research focuses on the task of Batched Vertex Cover Reconfiguration, both in centralized and distributed systems. In this talk, I will present a centralized black-box compression scheme for reconfiguration schedules. Afterwards, I will introduce the concept of Small Separator Decomposition which can be used to compute schedules in distributed systems and show how to compute it on specific graph classes in the LOCAL model of distributed computing. Lastly, I will complement the distributed results ...

[Full version]

[Full version]

Amazon Research, Alexa Shopping Internship Program Introduction

Thursday, 18.2.2021, 14:00

Zoom Event: Registration

Amazon Research, Alexa Shopping Internship Program Introduction on research challenges, and 2021 research internship program for graduate students in CS will be held on Thursday, February 18th between 14:00-15:00. Agenda: 14:00 - 14:20 “Alexa can you help me shop?“ Yoelle Maarek, VP of Research, Alexa Shopping, Amazon 14:20 - 14:30 Introduction to the 2021 internship program, Liane Lewin-Eytan, Sr Mgr., Alexa Shopping, Amazon 14:30 - 15:00 Panel & Q&As session, Moderated by Iftah Gamzu, Science ...

[Full version]

[Full version]

ceClub: Designing a Programming Language Shared-Memory Concurrency Semantics

A concurrency semantics (aka a memory model) for a programming language defines the allowed behaviors of multithreaded programs. For programmers, sequential consistency (i.e., standard interleaving-based semantics) is considered as the most intuitive model. However, it is too costly to implement. Designing a satisfactory substitute is highly challenging as it requires to carefully balance the conflicting desires of programmers, compilers, and hardware. In this talk I will introduce this challenge and the key ideas behind the ...

[Full version]

[Full version]

Pixel Club: Imaging with Local Speckle Intensity Correlations: Theory And Practice

Marina Alterman (EE, Technion)

Tuesday, 16.2.2021, 11:30

Zoom Lecture: https://technion.zoom.us/j/91594351204

Recent advances in computational imaging have significantly expanded our ability to image through scattering layers such as biological tissues, by exploiting the auto-correlation properties of captured speckle patterns. However, most experimental demonstrations of this capability focus on the far-field imaging setting, where obscured light sources are very far from the scattering layer. By contrast, medical imaging applications such as fluorescent imaging operate in the near-field imaging setting, where sources are inside the scattering layer. We ...

[Full version]

[Full version]

Heterogeneous Parametric Trivariate Fillets

Ramy Masalha

Monday, 15.2.2021, 13:30

Zoom Lecture:
6222766056

For password to lecture, please contact: sramy@cs.technion.ac.il

For password to lecture, please contact: sramy@cs.technion.ac.il

Blending and filleting are well established operations in solid modeling and computer-aided geometric design. The creation of a transition surface which smoothly connects the boundary surfaces of two (or more) objects has been extensively investigated. In this talk, we will introduce several algorithms for the construction of, possibly heterogeneous, trivariate fillets, that support smooth filleting operations between pairs of, possibly heterogeneous, input trivariates. A volumetric fillet, consisting of one or more tensor product trivariate(s), is ...

[Full version]

[Full version]

Characterizing, Exploiting, Detecting and Preventing DMA Attacks in the Presence of an IOMMU

Alex Markuze

Sunday, 14.2.2021, 15:00

Zoom Lecture:
99638464465

For password to lecture, please contact: markuze@cs.technion.ac.il

For password to lecture, please contact: markuze@cs.technion.ac.il

Malicious I/O devices might compromise the OS using DMAs. The OS therefore utilizes the IOMMU to map and unmap every target buffer right before and after its DMA is processed, thereby restricting DMAs to their designated locations. This usage model, however, is neither truly secure nor can it support multi-gigabit I/O operations. IOMMU provides protection at page granularity only, whereas DMA buffers can reside on the same page as other data leading to subpage vulnerabilities, ...

[Full version]

[Full version]

Indoor Exploration with a Robotic Vehicle Using a Single Camera and a Floorplan

John Noonan

Sunday, 14.2.2021, 12:00

Zoom Lecture:
2728213233

For password to lecture, please contact: John Noonan@cs.technion.ac.il

For password to lecture, please contact: John Noonan@cs.technion.ac.il

Intelligent systems which can be deployed to explore indoor buildings on a frequent and regular basis are beneficial to personnel operating remotely for security, manufacturing, or warehouse pack-and-ship. In this talk, I will present a new minimalistic approach to indoor exploration: minimal sensing, minimal prior map knowledge, and minimal underlying geometry needed to facilitate building a full visual scene representation. Our research combines both the classical and deep learning worlds, harnessing the strengths of each, ...

[Full version]

[Full version]

Data Science & Deep Learning: One-tape Turing Machine and Branching Program Lower Bounds for MCSP

Dimitrios Myrisiotis (Computing of Imperial College London)

Wednesday, 10.2.2021, 12:30

Zoom Lecture:
96255595054

For password to lecture, please contact: mayasidis@cs.technion.ac.il

For password to lecture, please contact: mayasidis@cs.technion.ac.il

For a size parameter s: N -> N, the Minimum Circuit Size Problem (denoted by MCSP[s(n)]) is the problem of deciding whether the minimum circuit size of a given function f: {0,1}^n -> {0,1} (represented by a string of length N := 2^n) is at most a threshold s(n). A recent line of work exhibited ``hardness magnification'' phenomena for MCSP: A very weak lower bound for MCSP implies a breakthrough result in complexity theory. For ...

[Full version]

[Full version]

Limited Associativity Caching in the Data Plane

Dor Hovav

Monday, 8.2.2021, 10:00

Zoom Lecture:
96832108498

For Password to lecture, please contact: dorhovav@cs.technion.ac.il

For Password to lecture, please contact: dorhovav@cs.technion.ac.il

In-network caching promises to improve the performance of distributed and networked applications. This is by storing so-called hot items in the network switches on-route between clients who need access to the data and the storage servers who maintain it. Since the data flows through those switches in any case, it is natural to cache hot items there. Programmable switches enable managing such caches in software, where the program gets compiled and then executed at ASIC ...

[Full version]

[Full version]

Approximating Requirement Cut via a Configuration LP

Yotam Sharoni

Sunday, 7.2.2021, 17:00

Zoom Lecture:
98726136846

For password to lecture, please contact: yotamsh@cs.technion.ac.il

For password to lecture, please contact: yotamsh@cs.technion.ac.il

We consider the REQUIREMENT CUT problem, where given an undirected graph G = (V, E) equipped with non-negative edge weights c , and g groups of vertices X1, . , Xg in V each equipped with a requirement ri, the goal is to find a collection of edges F in E, with total minimum weight, such that once F is removed from G in the resulting graph every Xi is broken into at least ri ...

[Full version]

[Full version]

CS Lecture: Deep into 3DV: Pushing the Boundaries of 3D Vision

Hadar Averbuch-Elor (Cornell-Tech)

Thursday, 4.2.2021, 16:30

Zoom Lecture:
98635528430

For password to lecture, please contact: sigal@cs.technion.ac.il

For password to lecture, please contact: sigal@cs.technion.ac.il

3D computer vision has significantly advanced over the past several decades, with modern algorithms successfully reconstructing entire urban cities. However, many questions remain unexplored, as geometric reasoning alone cannot fully infer the connections among images capturing different parts of the scene or semantic relationships between images captured at distant geographic locations. In this talk, I will present an ongoing line of research that leverages powerful deep networks to address new and exciting problems in 3D ...

[Full version]

[Full version]

CS Lecture: Towards Reliable Data-Driven Computations

Yuval Moskovitch (University of Michigan)

Monday, 1.2.2021, 16:00

Zoom Lecture:
97043323000

For password to lecture, please contact: sigal@cs.technion.ac.il

For password to lecture, please contact: sigal@cs.technion.ac.il

Data-driven methods are increasingly being used in domains such as fraud and risk detection, where data-driven algorithmic decision making may affect human life. The growing impact of data and data-driven systems on society makes it important that people be able to trust analytical results obtained from data-driven computations. This can be done in two complementary ways: by providing result explanations so that the user understands the computation and the basis for the observed results; and ...

[Full version]

[Full version]

ceClub: Comprehensive Protection for Speculatively-Accessed Data

Adam Morrison (Tel-Aviv University)

Wednesday, 27.1.2021, 11:30

Zoom Lecture: for link to zoom please contact sgoren@campus.technion.ac.il

Speculative execution attacks present an enormous security threat, capable of reading arbitrary program data under malicious speculation and later exfiltrating that data over microarchitectural covert channels. This talk will describe a comprehensive hardware protection from speculative execution attacks. We will first describe Speculative Taint Tracking (STT). STT delays the execution of instructions that create covert channels until their operands are proven to be a function of non-speculative data. STT builds on a comprehensive characterization of ...

[Full version]

[Full version]

CGGC Seminar: Errors in Judgement in Engineering: What Can They Teach Us about the Design Process?

Stefanie Elgeti (Institute of Lightweight Design and Structural Biomechnics,TU Wien)

Monday, 25.1.2021, 11:00

Zoom Lecture: 91344952941
For password to lecture please contact inbalb@cs.technion.ac.il

Engineering design is a task that comes with high responsibility: A failed design may easily cause not only monetary damage but, even more importantly, injuries of users. Based on a collection of design flaws [Petroski1994], this presentation will give an overview over modern design approaches that can help to prevent these mistakes in the future. It will touch upon both the topic of conceptual errors and numerical errors. The lecture will not be recorded.

[Full version]

[Full version]

Compositional Model Checking for Multi-Properties

Ohad Goudsmid

Sunday, 24.1.2021, 15:30

Zoom Lecture: for link to zoom please contact goudsmidohad@cs.technion.ac.il

Hyperproperties lift conventional trace properties in a way that describes how a system behaves in its entirety, and not just based on its individual traces. We generalize this notion to multi-properties, which describe the behavior of a set of systems, called a multi-model. We show that model-checking multi-properties is equivalent to model-checking hyperproperties. We introduce sound and complete compositional proof rules for model-checking multiproperties, based on approximations of the systems in the multi-model and describe ...

[Full version]

[Full version]

ceClub: Demand-Aware Optimization in Offchain Networks

Julia Khamis (EE, Technion)

Wednesday, 20.1.2021, 11:30

Zoom Lecture: for link to zoom please contact sgoren@campus.technion.ac.il

Offchain networks are dominant as a solution to the scalability problem of blockchain systems, allowing users to perform payments without their recording on the chain by relying on predefined payment channels. Users together with the offchain channels form a graph, known as the offchain network topology. A pair of users can employ a payment even without a direct channel through a path of channels involving other intermediate users. The offchain topology and payment characteristics affect ...

[Full version]

[Full version]

CS Lecture: Better Environments for Better AI

Sarah Keren (Harvard University and The Hebrew University of Jerusalem)

Tuesday, 19.1.2021, 10:30

Most AI research focuses exclusively on the AI agent itself, i.e., given some input, what are the improvements to the agent’s reasoning that will yield the best possible output? In my research, I take a novel approach to increasing the capabilities of AI agents via the use of AI to design the environments in which they are intended to act. My methods identify the inherent capabilities and limitations of AI agents and find the best ...

[Full version]

[Full version]

CGGC Seminar: Accelerating Geometric Algorithms for Freeform Surfaces using Toroidal Patch Approximation

We present a new approach to the acceleration of geometric algorithms for freeform surfaces using a hierarchy of bounding volumes, including those based on the osculating toroidal patches to the surfaces. Using this approach, we revisit some non-trivial conventional geometric algorithms, including those for computing the minimum and Hausdorff distances, the intersection and self-intersection curves, and the integral properties of freeform geometric models. We demonstrate the effectiveness of torus-based geometric computation, by reporting improvement in ...

[Full version]

[Full version]

CS Lecture: Learning on Pointclouds for 3D Scene Understanding

Or Litany (NVIDIA, Toronto AI lab)

Thursday, 14.1.2021, 17:00

Zoom Lecture:
91344952941

Meeting ID: 958 1720 7725 Passcode: CSLECTURE

Meeting ID: 958 1720 7725 Passcode: CSLECTURE

In this talk i'll be covering several works in the topic of 3D deep learning on pointclouds for scene understanding tasks. First, I'll describe VoteNet (ICCV 2019, best paper nomination): a method for object detection from 3D pointclouds input, inspired by the classical generalized Hough voting technique. I'll then explain how we integrated image information into the voting scheme to further boost 3D detection (ImVoteNet, CVPR 2020). In the second part of my talk I'll ...

[Full version]

[Full version]

Concurrent Sketches and their Applications

Dolev Adas

Thursday, 14.1.2021, 12:30

Sketches maintain compact approximate statistics about streams of data, thereby enabling quickly answering queries regarding the data stream without having to reprocess it. In this talk we will present four different papers that studies concurrent sketches and their applications. In particular we looked at these subjects : CRDT sliding window sketch, Multi-Producers Single-Consumer Queue, Limited Associativity Caches and Cache Admission Filter . In first result we introduce the notion of sliding window CRDT sketches where ...

[Full version]

[Full version]

CS Lecture: Computational Theory of Graphs, Sets and Rigid Sets

Nadav Dym (Duke University)

Tuesday, 12.1.2021, 16:00

Zoom Lecture:
91344952941

Meeting ID: 378 331 9350 Passcode: CSLECTURE

Meeting ID: 378 331 9350 Passcode: CSLECTURE

Quotient spaces are a natural mathematical tool to describe a variety of algorithmic problems where different objects are to be compared while their natural symmetries are to be ignored. In particular, we will focus on graphs and sets whose symmetries are permutation of the vertices, and rigid sets whose symmetries also include rigid motions. All three data types are prevalent in computer vision/graphics and in many other applications. We will discuss two problems involving these ...

[Full version]

[Full version]

CGGC Seminar: Quad-mesh Based Mappings between Surfaces

Helmut Pottmann (TU WIEN, Applied Geometry)

Monday, 11.1.2021, 11:00

Zoom Lecture: https://technion.zoom.us/j/91344952941

We discretize mappings between surfaces as correspondences between checkerboard patterns derived from quad meshes. This method captures the degrees of freedom inherent in smooth maps and provides a very simple and efficient computational approach to important types of maps such as conformal or isometric maps. In particular, it enables a natural definition of discrete developable surfaces which is much more flexible in applications than previous concepts of discrete developable surfaces. We discuss geometric modeling of ...

[Full version]

[Full version]

Campus Hour by ELBIT

Sunday, 10.1.2021, 17:30

Zoom Event: Registration

You are invited to a Campus Day by Elbit Systems, which presents opportunities and technologies, and a lecture by Yonatan Avraham, Development Team Leader, on unique solutions of infrastructure-free communication, as well as an open conversation with Guy Istmati, Director of Academy Relations, about career opportunities and recruitment processes. To view[Full version]

CS Lecture: Adversarially Robust Streaming Algorithms

Eylon Yogev (Tel-Aviv University & Boston University)

Thursday, 7.1.2021, 10:30

A streaming algorithm is given a long sequence of items and seeks to compute or approximate some function of this sequence using a small amount of memory. A body of work has been developed over the last two decades, resulting in optimal streaming algorithms for a wide range of problems. While these algorithms are well-studied, the vast majority of them are defined and analyzed in the static setting, where the stream is assumed to be ...

[Full version]

[Full version]

Pixel Club: Learned Sampling of 3D Point Clouds

Itai Lang (Tel-Aviv University)

Tuesday, 5.1.2021, 11:30

Zoom Lecture: https://technion.zoom.us/j/95495412165

There is a growing number of tasks that work directly on point clouds. As the size of the point cloud grows, so do the computational demands of these tasks. A possible solution is to sample the point cloud first. Classic sampling approaches, such as farthest point sampling (FPS), do not consider the downstream task. A recent work showed that learning a task-specific sampling can improve results significantly. However, the proposed technique did not deal with ...

[Full version]

[Full version]

CS Lecture: Next Generation Programming with Program Synthesis

Hila Peleg (CSE, University of California, San Diego)

Tuesday, 5.1.2021, 10:30

Program synthesis is the problem of generating a program to satisfy a specification of user intent. Since these specifications are usually partial, this means searching a space of candidate programs for one that exhibits the desired behavior. The lion's share of the work on program synthesis focuses on new ways to perform the search, but hardly any of this research effort has found its way into the hands of users. We wish to use synthesis ...

[Full version]

[Full version]

CGGC Seminar: Hyper-Realistic Rendering: Leveraging Artistic & Mathematical Approaches for Effective Control of Visual Results

Ergun Akleman (Texas A&M University)

Monday, 4.1.2021, 16:00

Zoom Lecture: https://technion.zoom.us/j/91344952941

My primary goal in this fringe direction of research is to develop a simple, intuitive formal framework for the automatic representation of simplified shapes and materials that can support Hyper-Realism in a wide variety of rendering applications. I observe that with the emphasis on the physical laws in rendering systems, (1) the focus increasingly shifts away from how users perceive the virtual environment, (2) rendering becomes prohibitively difficult to realize desired global illumination effects in ...

[Full version]

[Full version]