SIPL Annual Event 2019 (July 1st, 2019)

Gathering & Poster Session




Prof. David Malah, Head of SIPL

Invited Talk: OttoFence – Automotive Cyber Security


Naftaly Sharir

Modern vehicles of today rely on CANBUS networks for all their critical systems. CANBUS is a shared bus architecture. A single misbehaving node can effectively block all network communication say by generating a Denial-Of-Service attack. As most of the cars now are connected cars, this situation may end up with a successful ransomware cyber-attack, up to life threatening attacks. What can be done to prevent this? What is done today and what are the plans for next…. ? I will explore possible hardware based Automotive Stateful Firewall solution, and will short demo the situation without and with the proposed solution.

Naftaly Sharir is co-founder, CEO & CTO at OttoFence. He is a serial entrepreneur, technical & business leader and manager: HW , SW and systems. He has 30 years of business and technical experience in communication, internet, DSP, audio, video, VLSI and cyber. His reach career in multimedia signal processing, wireless and mobile, began at SIPL and IBM Haifa Research Labs. Following as VP R&D at VDOnet, CEO at Emblaze Semi, CEO at Electronics-Line 3000 Ltd, CEO at Advasense, co-founder at Vitalitix, co-founder at Pixie-Technology and CTO at Terafence.

Wilk Family Awards and Outstanding Supervisor Awards Ceremony


Domain Adaptation Using Riemannian Geometry of SPD Matrices


Prize winner in the Kasher undergraduate project contest in the Faculty of Electrical Engineering

Gal Maman

Supervisor: Or Yair

In collaboration with Prof. Ronen Talmon & Dr. Danny Eytan (Rambam Medical Center)

Presented at ICASSP 2019

Break & Poster Session


Review of Teaching Activity in SIPL


Nimrod Peleg

Robust Automatic Detector and Feature Extractor for Dolphin Whistles


Wilk family award winner

Yoel Bud, Guy Shkury

Sueprvisor: Dr. Roee Diamant (School of Marine Sciences, University of Haifa)

Presented at OCEANS 2019

Audio-visual Processing of Speech with Deep Neural Networks


Ido Ariav, M.Sc. student

Advisors: Prof. Israel Cohen

Published in:

  • Signal Processing, January 2019
  • IEEE Journal on Selected Topics in Signal Processing, May 2019


Recently, there has been growing use of deep neural networks (DNN) in many modern speech-based systems such as speaker recognition, speech enhancement, and emotion recognition. Inspired by this success, we propose to address the task of voice activity detection by incorporating auditory and visual modalities into a DNN. In this talk, I will survey several classic voice activity detection methods and present two novel multimodal deep architectures for this task. I will demonstrate the advantages of such deep architectures in challenging acoustic environments including high levels of noise and transients, which are common in real life scenarios.

SinGan: Learning a Generative Model from a Single Natural Image


Tamar Rott Shaham, Ph.D. student

Advisor: Prof. Tomer Michaeli

Submitted to ICCV 2019


We introduce SinGAN, an unconditional generative model that can be learned from a single natural image. Our model is trained to capture the internal distribution of patches within the image, and is then able to generate high quality, diverse samples that carry the same visual content as the image. SinGAN contains a pyramid of fully convolutional GANs, each responsible for learning the patch distribution at a different scale of the image. This allows generating new samples of arbitrary size and aspect ratio, that have significant variability, yet maintain both the global structure and the fine textures of the training image. In contrast to previous single image GAN schemes, our approach is not limited to texture images, and is not conditional (i.e. it generates samples from noise). User studies confirm that the generated samples are commonly confused to be real images. We illustrate the utility of SinGAN in a wide range of image manipulation tasks.

Live Demos

  • Manifold learning for data-driven dynamical system modeling, Kobi Shiran & Gal Kinberg, supervisor: Or Yair, presented at ICASSP 2019
  • Android app for pedestrian traffic light recognition, Roni Ash & Dolev Ofri, supervisor: Yair Moshe, in cooperation with the Technion’s Social Hub
  • Physics classroom augmented reality with your smartphone, Yonahan Sackstein & Georgee Tsintsadze, supervisor: Yair Moshe
  • TAKI cards game judge, Shai Lottem & David Shwartsburd, supervisor: Ori Bryt
  • Ultrasonic 3D positioning system with unsynchornized beacons and receivers, Guy Dascalu & Omer Movshovits, supervisor: Alon Eilam
  • Proximity sensor for smartphones based on acoustic measurements, Andy Rodan & Zacharie Cohen, supervisor: Pavel Lifshits, in cooperation with DSP-Group
  • Protractor for smartphones based on acoustic measurements, Shay Avig & Matanel Yaacov, supervisor: Alon Eilam
  • Speech2singing, Yair Yarden & Ofir Kedem, supervisor: Yair Moshe, in cooperation with Elad Keidan