Overview
Explore image restoration using deep learning techniques, focusing on dehazing, in this comprehensive talk by Subhaditya Mukherjee, a computer science student and ML facilitator. Discover how neural networks outperform classical methods in restoring extremely hazy images, learn about the power of deep learning in image restoration tasks, and gain insights into specific techniques such as Units, Dynamic Units, and segmentation. Delve into the architecture and applications of these methods, and understand their potential for generalizing across various restoration tasks.
Syllabus
Introduction
Agenda
Overview
Dehazing
Units
Why Units
Unit Architecture
Dynamic Units
Reshoot
Random image
Small image
What is segmentation
Part from Shadows
Unit
More papers
Questions
Taught by
Abhishek Thakur