Explore a comprehensive IEEE Signal Processing Society lecture on signal processing advancements, from compressed sensing to deep learning. Delve into task-based structured acquisition, sub-Nyquist sampling techniques, and cognitive radar applications. Discover innovative approaches in ultrasound imaging, super-resolution microscopy, and analog-to-digital compression. Examine the unification of rate-distortion and sampling theory, and investigate the interplay between model-based signal processing and deep learning. Learn about cutting-edge algorithms like SPARCOM, SUSHI, and DUBLID, and their applications in various fields. Gain insights into data-driven hybrid algorithms and factor graph methods, bridging traditional signal processing with modern machine learning techniques.
From Compressed Sensing to Deep Learning - Tasks, Structures and Models
IEEE Signal Processing Society via YouTube
Overview
Syllabus
Intro
Data Redundancy
Digital Information
Analog Girl in a Digital World...
Standard Acquisition Systems
Limitations of Standard Systems
Task-Based Structured Acquisition
Advantages of Joint Design
Streams of Pulses Radar
Xampling Hardware
Compressed Sensing Extensions
Sub-Nyquist Ultrasound Imaging
Demo Movie
Deep Adaptive Beamforming
Channel Data Clinical Forum Improve diagnostics from channel data!
Sub-Nyquist and Cognitive Radar
Cognitive Automotive Radar
Multicoset Sampling
Xampling: Modulated Wideband Converter
Sub-Nyquist Cognitive Radio
Super Resolution Microscopy
SPARCOM: Super Resolution Correlation Microscopy
Super Resolution Contrast Enhanced Ultrasound
SUSHI: Sparsity-Based Ultrasound Super- resolution Hemodynamic Imaging
Analog to Digital Compression
Unification of Rate-Distortion and Sampling Theory
Quantizing the Samples: Source Coding Perspective
Optimal Sampling Rate
Metasurfaces for Analog Precoding
Antenna Selection for Imaging
Product Arrays
Spatial Sub-Sampling
Black-Box Deep Learning
Model Based Signal Processing
Model-Based vs. Deep Learning Model-based signal processing
Model-Based Deep Learning
Deep Unfolding
DUBLID: Deep Unrolling for Blind Deblurring
Deblurring Results
Super-resolution via Deep Learning
Data Driven Hybrid Algorithms
Data-Driven Factor Graph Methods
Taught by
IEEE Signal Processing Society
Related Courses
-
Model Based Deep Learning - Applications to Imaging and Communications - Yonina Eldar
-
Computational Imaging Systems: From DiffuserCam to Neural Activity Tracking - Seminar 2
-
Model Based Deep Learning with Application to Super Resolution - IPAM at UCLA
-
Computational Microscopy Using LED Array for 3D Phase Imaging - Prof. Lei Tian
