YoVDO

Set Distribution Networks - A Generative Model for Sets of Images

Offered By: Yannic Kilcher via YouTube

Tags

Generative Models Courses Machine Learning Courses Computer Vision Courses Variational Methods Courses Image Generation Courses

Course Description

Overview

Explore a comprehensive video explanation of the paper "Set Distribution Networks: a Generative Model for Sets of Images". Delve into the novel approach of generating sets of images with shared characteristics, particularly when the number of sets is unknown and potentially infinite. Learn about the probabilistic framework and practical implementation of this generative model based on variational methods. Understand the architecture, including the set encoder, discriminator, generator, and prior. Follow the detailed breakdown of the model's components, likelihood function, loss function, and optimization techniques. Examine the results and potential applications in face verification and 3D reconstruction. Gain insights into this cutting-edge research that pushes the boundaries of generative modeling for image sets.

Syllabus

- Intro & Overview
- Problem Statement
- Architecture Overview
- Probabilistic Model
- Likelihood Function
- Model Architectures
- Loss Function & Optimization
- Results
- Conclusion


Taught by

Yannic Kilcher

Related Courses

6.S191: Introduction to Deep Learning
Massachusetts Institute of Technology via Independent
Generate Synthetic Images with DCGANs in Keras
Coursera Project Network via Coursera
Image Compression and Generation using Variational Autoencoders in Python
Coursera Project Network via Coursera
Build Basic Generative Adversarial Networks (GANs)
DeepLearning.AI via Coursera
Apply Generative Adversarial Networks (GANs)
DeepLearning.AI via Coursera