SIPL Annual Event 2018 (July 2nd, 2018)

Photos icon Photos from the Event


Gathering & Poster Session

13:30-14:00


Video icon Welcome

14:00-14:10

Prof. David Malah, Head of SIPL


 Video icon PDF Logo Improving Training Efficiency in Deep Learning

14:10-14:35

Prof. Daniel Soudry

Abstract:

I will present several empirical and theoretical results related to improving training efficiency in deep networks, based on the following works:


Video icon Wilk Family Awards and Outstanding Supervisor Awards Ceremony

14:35-14:45


 Video icon PDF Logo Unsynchronized Acoustic Indoor Positioning

14:45-15:05

2nd prize winner in the Kasher undergraduate project contest in the Faculty of Electrical Engineering

Guy Feferman, Michal Blatt

Supervisor: Alon Eilam

In cooperation with:  Sonarax Logo

ICASSP 2018 Demo


Break & Poster Session

15:05-15:35


 Video icon PDF Logo Review of Teaching Activity in SIPL

15:35-15:50

Nimrod Peleg


Video icon PDF Logo Detection and Localization of Cumulonimbus Clouds in Satellite Images

15:50-16:10

Wilk family award winner

Etai Wagner, Ron Dorfman

Supervisor: Almog Lahav

In cooperation with:  Rafael Logo

Submitted to ICSEE 2018


Video icon PDF Logo Local-to-Global Point Cloud Registration using a Viewpoint Dictionary

16:10-16:35

David Avidar, M.Sc. student

Advisors: Prof. David Malah, Dr. Meir Bar-Zohar

Partly funded by the OMEK consortium

Presented at ICCV 2017

Abstract:

Local-to global point cloud registration is a challenging task due to the substantial differences between these two types of data, and the different techniques used to acquire them. Global clouds cover large-scale environments and are usually acquired aerially (e.g., using Airborne Laser Scanning – ALS), and local clouds are often acquired from ground level at a much smaller range (e.g., using Terrestrial Laser Scanning – TLS). As a result of these differences, existing point cloud registration approaches, such as keypoint-based registration, tend to fail.

We propose a novel registration method based on converting the global cloud into a viewpoint-based dictionary. We associate each viewpoint with a panoramic range-image, capturing the geometry of the visible environment. Then, plausible local-to-global transformations can be found via a dictionary search. We show efficient dictionary search can be done using phase-correlation between panoramic range-images.

We demonstrate that the proposed viewpoint-dictionary-based registration method achieves better performance than state-of-the-art, keypoint-based methods (e.g., FPFH, RoPS), even without any GPS measurements. For the evaluation, we used a challenging dataset of 108 TLS local clouds and an ALS large-scale global cloud, in a 1km2 urban environment.


Video icon PDF Logo The Perception-Distortion Tradeoff

16:35-17:00

Yochai Blau, Ph.D. student

Advisor: Prof. Tomer Michaeli

Presented at CVPR 2018

Abstract:

Image restoration algorithms are typically evaluated by some distortion measure (e.g. PSNR, SSIM, IFC, VIF) or by human opinion scores that quantify perceived perceptual quality. In this work, we prove mathematically that distortion and perceptual quality are at odds with each other. Specifically, we study the optimal probability for correctly discriminating the outputs of an image restoration algorithm from real images. We show that as the mean distortion decreases, this probability must increase (indicating worse perceptual quality). As opposed to the common belief, this result holds true for any distortion measure, and is not only a problem of the PSNR or SSIM criteria. However, as we show experimentally, for some measures it is less severe (e.g. distance between VGG features). We also show that generative-adversarial-nets (GANs) provide a principled way to approach the perception-distortion bound. This constitutes theoretical support to their observed success in low-level vision tasks. Based on our analysis, we propose a new methodology for evaluating image restoration methods, and use it to perform an extensive comparison between recent super-resolution algorithms. Our study reveals which methods are currently closest to the theoretical perception-distortion bound.

 


Posters

Project TitleStudentsSupervisorComments / Collaboration
Acoustic positioning with unsynchronized sound sourcesGuy Feferman
Michal Blatt
Alon EilamGuy Shofen
Sonarax
Influence of Motion on Video Quality Perception
Ido Lev and Ilan MeltzerYair MosheHarmonic
General Road Detection From A Single Image
Dudi LubtonYair Moshe
Interference Cancellation for Underwater Acoustic Signals
Eitan Ovrutski
Sapir Madi
Dr. Roee Diamant
Peripheral Vascular Disease -
Walking Pattern Recognition

Sztyglic Ori
Septon Shay
Ofer DaninoProf. Aharon Hoffman (Rambam hospital)
Android Application for Audio QRRon Steinberg
Zvi Herscovici
Alon EilamSonarax
Detection and Localization of Cumulonimbus Clouds in Satellite Images
Ron Dorfman and Etai WagnerAlmog LahavWilk Award 2018,
RAFAEL
Audio Retrieval by Voice Imitation
Samah Khawaled Mohamad KhatibDr. Hadas BenistyWilk Award 2018
Monitoring Vital Signs of Children in Intensive Care
Ifat Abramovich and Tomer Ben YehudaOr YairDr. Danny Eytan (Rambam hospital)
Estimating Brain Activation Patterns from EEG Data
Ronen Rahamim and Gilad HechtOr YairDr. Danny Eytan (Rambam hospital)
Differential Microphone Arrays (DMA)
Demo System
Ronen SheffiKobi Bucris and Dr. Evyatar Hemo
RAFAEL
Semi-Supervised Classification of Acoustic SignalsInbar Rachel FrenkelRon Amit
RAFAEL
Real-Estate market estimation
Using Machine Learning algorithms
Roei Shemesh and Nitzan Krasney Jonatan Chernyak
Detect and Classify Species of Fish
Shalom Lagziel
Idan Kligvasser
Teaching Deep Networks to Improvise Jazz
Yamin Ben Shimon Shunit Haviv
Nadav Bhonker
Revealing ADHD from functional MRIYossi Levi
Itzik Merkado
Supervised by Andrey Zhitnikov
Prof. Tzipi Horowitz-Kraus, Technion
Reading Difficulty Detection using Eye MovementsJonathan Shani
Yoni Osin
Or Yair
Prof. Tzipi Horowitz-Kraus
News Stories Classification Using Online Deep LearningInna Batenkov
Yoni Lederman
Yaron EzraRAFAEL
Classification of dolphin's whistles using Deep LearningMatan Tur-paz
Shahaf Aricha
Yaron Doweck and
Dr. Alon Amar
RAFAEL
Recognition of PVD based on Walking Pattern Acquired from SmartphoneOri Cohen
Samyon Shahnovitch
Supervised by Ofer Danino and Erez Fridman
Prof. Aaaron Hoffman
Speech Enhancement Using Deep LearningAriel Zelman
Tzvika Sweetkin
Yohai Blau and Dr. David Dov
Elbit Systems
Fast Video Encoding using Deep LearningOmri Carmi and Harel HaskyYair MosheHarmonic
In progress
Alarm Sound Detection Using Deep Learning
Yoni Lederman and Naor FadidaYair MosheIn progress
Towards Blood Pressure Estimation Using a Smartphone CameraIdo Lev and Inbal RonaiYair MosheIn progress
Identification Of Content In Images Of Promotional BrochuresGil Gelbert
Ilan Meltzer
Ori BrytQuantaCom
In progress
Clustering of Sounds in Humpback Whale SingingGal Harush
Inbal Kom-Betzer
Dr. Alon Amar and Yaron Doweck
RAFAEL
In progress
Artifact Reduction From Brain Activity SignalsBarak Hoffer
Einav Aharon
Dr. David DovAlpha-Omega
In progress
Examination of TMS Effect Using EEG SignalMatan Allouche Reggev LivneyOr YairBrainsWay
In progress
Video Classification Using Deep LearningIfat Abramovich Tomer Ben YehudaDr. Rami CohenIn progress
Efficient Deep Learning for
Pedestrian Traffic Light Recognition
Roni Ash
Dolev Ofri
Yair MosheIn progress
Heart Disease Detection from Human VoiceBen Dgani
Eli Szulc
Supervised by Kfir Shem-TovCardioKol
In progress
Gesture Recognition With a Smart BraceletAndre Kassis
Majd Abu Tayeh
Ariel Schwartz
In progress
Multi-Microphone Speaker Localization Demo SystemAvi Semo
Alon Saguy
Kobi Bucris and Dr. Evyatar Hemo
In progress
Anomaly Detection on Chest X-Rays
with Deep Learning
Yair Yarden
Ofir Kedem
Yair MosheIn progress
Deep Learning for Classroom Augmented Reality ApplicationTom Kratter
Yonatan Sackstein
Yair MosheIn progress
Color Calibration For Smart Product LabelsGil GelbertOri BrytFreshPoint
In progress
Detection and Analysis of Dolphin SoundsGuy Shkury
Yoel Bud
Dr. Roee DiamantIn progress
Signal Processing Algorithms Development and Implementation with Android DSP LabMatanel Yaacov
Shay Avig
Alon EilamIn progress
Android App for Detecting Alarm SoundsRachel Gringols Michal Kuperman Koren Alon Eilam and Yair Moshe
In progress
Synthesis of Mammography Images Using Deep LearningLotem Manassen
Inbar Kazes
Yochai Blau and Idan KligvasserIBM
In progress
Acoustic Beamformer with a Constant Beamwidth for Planar ArrayRoni Ash
Arkadi Gurevich
Dr. Baruch Berdugo
Israel Police Dept.
In progress
Estimating Brain Activation Patterns from EEG DataGal Maman Or Yair and Prof. Ronen TalmonDr. Danny Eitan
In progress
Cardiac Arrest Prediction Based on Geometric Modeling of Pediatric CCU Monitor DataGal Shachaf Tal Shnitzer and Prof. Ronen TalmonDr. Danny Eytan
In progress
Learning to Deform for Efficient Image CompressionElad Hirsch
Nadav Eliyahu
Tamar Rott and Idan Kligvasser
In progress
Wind Noise Reduction from Speech SignalsAdir Kobovich
Tomer Lerner
Tal ShnitzerWaves
In progress
Relative Camera Pose Estimation Using Smartphone Cameras and SensorsOmer Movshovitz
Guy Dascalu
Ori Bryt and Adam Wolff
RAFAEL
In progress
Stitching Corrections For Aerial Images Using Deep LearningTomer Rosenbaum and Dror Kabely
Ori BrytRAFAEL
In progress
Residual Echo Suppression Using Deep LearningAya Ganem
Omer London
Andrey Zhitnikov
In progress
Semi-supervised Learning for Acoustic Source Localization and TrackingAvner Okun
Barak Gerstein
Ori KatsIn progress
Android App for Detecting Alarm SoundsRachel Gringols Michal Kuperman KorenAlon Eilam and Yair Moshe
In progress
Obtaining Physical Laws Empirically with Manifold LearningKobi Shiran
Gal Kinberg
Or YairMr. Ofer Danino

In progress
Interactive Demo Systemfor 3D Depth Data ProcessingLital Filber and Ori Lador,Yair MosheIn progress
From Deep Features To Image RestorationOri SztyglicTamar Rott Shaham and Idan KligvasserIn progress
Image Compression Using Deep Learning TechniquesIdan Berkovich
Omer Gal
Supervised by Ori Bryt, Nadav Bhonker, Dr. Ron Rubinstein
In progress
Real-time Audio Processing for AndroidIvan Zorin
Alex Dashevski
Nimrod Peleg and Pavel LifshitsIn progress