(The talk will be given in English)

Speaker:     Dr. Assaf Shocher

                              NVIDIA

Monday, December 23^rd, 2024

12:00 - 13:00

Making Neural Networks Linear Again: Projection and Beyond

Abstract

Every day, somewhere, a researcher mutters, “If only neural networks were linear, this problem would be solved”. Linear operations offer powerful tools: projection onto subspaces, eigen decomposition, and more. This talk explores their equivalents in the non-linear world of neural networks, with a special focus on projection, generalized by idempotent operators- operators that satisfy f(f(x)) = f(x).

Idempotent Generative Network (IGN) is a generative model that is trained by enforcing two main objectives: (1) target distribution data map to themselves f(x) = x, defining the target manifold, and (2) latents project onto this manifold via the idempotence condition f(f(z)) = f(z). IGN generates data in a single step, but can iteratively refine, and projects corrupted data back onto the distribution.

This projection ability gives rise to Idempotent Test-Time Training (IT³), a method to adapt models at test time using only current out-of-distribution (OOD) input. During training, the model f receives an input x along with either the ground truth label y or a neutral "don't know" signal ∅. At test-time, given corrupted/OOD input x, a brief training session minimizes ||f(x, f(x, ∅)) - f(x, ∅)||, making f(x,⋅) idempotent. IT³ works across architectures and tasks, demonstrated for MLPs, CNNs, and GNNs on corrupted images, tabular data, OOD facial age prediction, and aerodynamic predictions.

Finally, I'll ask: "Who says neural networks are non-linear?" They're only non-linear with respect to the standard vector spaces! In an ongoing work, we construct vector spaces X, Y with their own addition, negation, and scalar multiplication, where f: X → Y becomes truly linear. This enables novel applications including spectral decomposition, zero-shot solutions to non-linear inverse problems via Pseudo-Inverse, and architecture-enforced idempotence.

Short Bio

I am a postdoctoral researcher at NVIDIA. Prior to that I was a postdoctoral fellow at UC Berkeley, working with Alyosha Efros, and a visiting researcher at Google. I received my PhD from the Weizmann Institute of Science, where I was advised by Michal Irani. I have bachelor's degrees in Physics and EE from Ben-Gurion University. My prizes and honors include the Rothschild postdoctoral fellowship, the Fulbright postdoctoral fellowship, John F. Kennedy award for outstanding Ph.D. at the Weizmann Institute, and the Blavatnik award for CS Ph.D. graduates.

השתתפות בסמינר תיתן קרדיט שמיעה = עפ"י רישום שם מלא + מספר ת.ז. בטופס הנוכחות שיועבר באולם במהלך הסמינר

:Abstract

Histology and Cytology play a critical role in biomedical research by enabling the detection and analysis of tissue and cell- Morphology, abnormalities, protein and RNA expression and treatments effects. Over the past decade, significant advancements have transformed this field, including the development of synthetic biomarkers and new methods that enable the use of multiple fluorescence-labeled antibodies and RNA probes on single sample section. Additionally, innovations in confocal microscopy and high-resolution slides scanning microscopes have greatly improved imaging capabilities. The emergence of Machine learning based Quantitative Image Analysis- that enables precise analysis of histology and cytology large datasets (i.e. cell populations, distances between cell populations, expression level in each cell population), patterns recognition (i.e. blood vessels populations detection and analysis, neuronal and collagen fibers parameters analysis etc.).
Together- these advancements in Histology and cytology, microscopy and image analysis- enable researchers to extract high quantity of data from each sample section, enabling accurate quantitative evaluation of basic research data and treatments effects.
In this seminar, I will present the main Histology, Cytology and Machine learning based Quantitative Image Analysis (HCA) methodologies and innovations in relevant research fields, providing examples in different cells, tissues and animal models. This will equip researchers with a better understanding of .how to apply these cutting-edge HCA techniques to their own .studies 

  -סמינר זה יחשב כסמינר שמיעה לתלמידי תואר שני ושלישי-  This Seminar Is Considered A Hearing Seminar For Msc/Phd Students

Direct and Inverse Problems in Ocean Acoustics: Propagation, Physics-Based Processing and Applications

Monday February 3th 2025 at 14:00 

Wolfson Building of Mechanical Engineering, Room 206 

Abstract:

The undersea world presents a fascinating yet challenging environment for sensing, with acoustics serving as the primary tool due to the limited propagation of electromagnetic waves in seawater. Sound, however, can travel vast distances underwater, sometimes propagating thousands of kilometers under specific conditions. Yet, the ocean’s highly inhomogeneous nature, characterized by strong spatial and temporal variability, creates a highly complex acoustic environment, posing significant challenges for both direct and inverse problems in ocean acoustics.

This seminar explores these challenges by addressing one direct and one inverse problem, each in drastically different acoustic environments. First, we investigate how internal Kelvin gravity waves impact sound propagation in shallow water. Through analysis of experimental data from Lake Kinneret and advanced sound propagation modeling, we demonstrate that internal Kelvin waves can cause a significant, fourfold decrease in sound intensity over a 5-km acoustic track — an important consideration for underwater communication and defense applications.

The second part of the seminar moves to the Challenger Deep, the Earth’s deepest point, where we employ a synthetic aperture autonomous recording system to address an inverse problem. Here, we demonstrate how wind-driven ambient noise can be leveraged to infer ocean acidity — a critical concern due to its impact on marine ecosystems. By analyzing the depth-dependent attenuation of sound and applying the principles of passive absorption spectroscopy, we showcase a novel method for estimating the volumeintegrated pH of seawater

Bio:

Ernst began his academic journey studying Nuclear Physics and Technology in Russia before making Aliyah and pursuing his Master’s and PhD studies in Underwater Acoustics in Haifa University under the supervision of Prof. Boris Katsnelson, graduating with honors. His research involved extensive experimental work in Lake Kinneret and Mediterranean Sea, focusing on sound propagation in shallow and deep water affected by spatial and temporal inhomogeneities. At the latest stage of his PhD, Ernst received the prestigious Batsheva de Rothschild scientific grant and served as a Visiting Researcher at the Naval Postgraduate School, where he studied single-element time-reversal mirrors.

He later transitioned to nonlinear acoustics, conducting postdoctoral research on parametric acoustic and receiving arrays at the Technion’s Department of Mechanical Engineering under supervision of Prof. Izhak Bucher. Following this, Ernst undertook another postdoctoral position at Dalhousie University in Canada, developing deep-ocean noise spectroscopy methods using data from an autonomous synthetic aperture ocean profiler under the supervision of Dr. David Barclay.

Currently, Ernst is conducting postdoctoral research at the Naval Postgraduate School in the United States under supervision of Prof. Oleg Godin, supported by a grant from the National Academy of Sciences. He is actively involved in the New England Seamounts Acoustic Experiment, investigating acoustic propagation and associated diffraction and refraction effects around prominent bathymetric features.

Ernst is a member of the Acoustical Society of America and its Technical Committee on Acoustical Oceanography. He has chaired sessions at Acoustical Society of America meetings and serves as a reviewer for leading journals, including the Journal of the Acoustical Society of America, Applied Acoustics and the Journal of Sound and Vibration and.