Description
Venue
Online only
Period
Two days
Capacity
Five students
Qualification
Programming skill in C/C++ or Python.
Summary
In recent years, deep learning approaches have achieved remarkable success in image processing. On the other hand, the performance of deep learning depends on the size and quality of training dataset. This seminar focuses on image defogging, and we evaluate the performance and limitation of deep learning through synthesizing foggy images with a physics-based model and training a deep neural network.