Communications and Signal Processing Seminar Series

1.62K subscribers

127 videos

This is the channel for the Communications and Signal Processing (CSP) Seminar Series hosted by Electrical and Computer Engineering at the University of Mich...

Latest Videos

Inference-Time Information Theory

Towards Secure and Safe AI Agents

Tail Distribution of Regret in Optimistic Reinforcement Learning

Signal, Image Processing, and Machine Learning for Reticle Inspection.

Input-dependent Fisher information for interpreting deep classifiers and identifying task-relevant

Generalization Properties of Score-matching Diffusion Models for Intrinsically Low-dimensional Data

Inter-temporal Incentive and Its Application in Strategic Classification and Voluntary Audit

Towards AI-assisted Video Editing Generating Shorts from Long Videos

Large-scale computational imaging with highly parallelized, camera array-based detection

How do LLMs generalize on out-of-distribution tasks insights from model’s internal representations

“Robust and Geometry-Aware Distribution Estimation via Optimal Transport”

“Consumer Profiling via Information Design” by Itay P. Fainmesser, Andrea Galeotti, and Ruslan Momot

Advancing Efficient and Trustworthy AI for Science, Engineering, and Medicine

From Generative AI to Enhanced Dementia Care The Path to Early Detection and Intervention

The Interplay between Privacy and Robustness from Mean Estimation to LLM Alignment

Unsourced Random Access, Approximated Message Passing, and Cell-Free Wireless Networks.

Guiding Diffusion and Flow Models for Constrained Sampling in Image, Video and 4D

Learning to detect an anomalous Markov process

Behavior of Gradient-Based Optimization Methods in Nonconvex Landscapes A Dynamical Systems Perspec

Setup Times in Multiserver Systems New Insights

Machine Unlearning for Generative AI

Exploring and Mitigating Safety Risks in Large Language Models and Generative AI

ADVANCING IMAGE RECONSTRUCTION AND RESTORATION THROUGH ROBUST SUPERVISED AND GENERATIVE MODELS

Improved very sparse matrix completing using an intentionally randomized “asymmetric SVD”

Enhancing Fairness in Deep Learning from Data and Model Perspectives

Design-Based Confidence Sequences A General Approach to Risk Mitigation in Online Experimentation

Mixed Reality Security and Streaming at the System and Application Levels

How to make the gradient descent-ascent converge to local minimax optima

Transformer Meets Nonparametric Kernel Regression

Recent Advances in Average-Reward Restless Bandits