You are here

Upcoming Seminars

MondayNov 19, 201814:30
Foundations of Computer Science SeminarRoom 155
Speaker:Natan RubinTitle:Hitting Convex Sets with PointsAbstract:opens in new windowin html    pdfopens in new window

Given an underlying finite point set P in the plane, we seek a small set Q that would hit any convex set that contains at least an Epsilon-fraction of P. Such a set Q is called a weak Epsilon-net. The study of Epsilon-nets is central to Computational and Combinatorial Geometry, and it bears important connections to Statistical Learning Theory, Extremal Combinatorics, Discrete Optimization, and other areas.

It is an outstanding open problem to determine tight asymptotic bounds on weak Epsilon-nets with respect to convex sets. For any underlying point set in the plane we describe such a net whose cardinality is roughly proportional to Epsilon^{-3/2}. This is the first improvement of the over-25-year-old bound of Alon, Barany, Furedi, and Kleitman.

TuesdayNov 20, 201809:15
Computer Science Seminar
Speaker:Moshe TennenholtzTitle:Data Science with Game Theory FlavorAbstract:opens in new windowin html    pdfopens in new windowBenozio Building, Room 691

Design of data science algorithms and techniques, central to the Internet and on-line media, needs to be revolutionized. Current designs ignore participants' strategic incentives. We are establishing an entirely new repertoire of incentive-compatible data science algorithms and techniques, with major applications in search and information retrieval, recommendation systems, regression, on-line learning, clustering and segmentation, and social networks analysis. In this talk I will introduce our research agenda, and discuss in more detail couple of concrete contributions


ThursdayNov 22, 201812:15
Vision and Robotics SeminarRoom 1
Speaker:Yehuda Dar Title:System-Aware Compression: Optimizing Imaging Systems from the Compression StandpointAbstract:opens in new windowin html    pdfopens in new window

In typical imaging systems, an image/video is first acquired, then compressed for transmission or storage, and eventually presented to human observers using different and often imperfect display devices. While the resulting quality of the perceived output image may severely be affected by the acquisition and display processes, these degradations are usually ignored in the compression stage, leading to an overall sub-optimal system performance. In this work we propose a compression methodology to optimize the system's end-to-end reconstruction error with respect to the compression bit-cost. Using the alternating direction method of multipliers (ADMM) technique, we show that the design of the new globally-optimized compression reduces to a standard compression of a "system adjusted" signal. Essentially, we propose a new practical framework for the information-theoretic problem of remote source coding. The main ideas of our method are further explained using rate-distortion theory for Gaussian signals. We experimentally demonstrate our framework for image and video compression using the state-of-the-art HEVC standard, adjusted to several system layouts including acquisition and rendering models. The experiments established our method as the best approach for optimizing the system performance at high bit-rates from the compression standpoint.
In addition, we relate the proposed approach also to signal restoration using complexity regularization, where the likelihood of candidate solutions is evaluated based on their compression bit-costs.
Using our ADMM-based approach, we present new restoration methods relying on repeated applications of standard compression techniques. Thus, we restore signals by leveraging state-of-the-art models designed for compression. The presented experiments show good results for image deblurring and inpainting using the JPEG2000 and HEVC compression standards.
* Joint work with Prof. Alfred Bruckstein and Prof. Michael Elad.
** More details about the speaker and his research work are available at



ThursdayNov 29, 201812:15
Vision and Robotics SeminarRoom 155
Speaker:Yair Weiss Title:Why do deep convolutional networks generalize so poorly to small image transformations?Abstract:opens in new windowin html    pdfopens in new window

Deep convolutional network architectures are often assumed to guarantee generalization for small image translations and deformations. In this paper we show that modern CNNs (VGG16, ResNet50, and InceptionResNetV2) can drastically change their output when an image is translated in the image plane by a few pixels, and that this failure of generalization also happens with other realistic small image transformations. Furthermore,  we see these failures to generalize more frequently in more modern networks. We show that these failures are related to the fact that the architecture of modern CNNs ignores the classical sampling theorem so that generalization is not guaranteed. We also show that biases in the statistics of commonly used image datasets makes it unlikely that CNNs will learn to be invariant to these transformations. Taken together our results suggest that the performance of CNNs in object recognition falls far short of the generalization capabilities of humans.
Joint work with Aharon Azulay