Deep Reinforcement Learning for Object Tracking

Overview

This course covers the learning outcomes and goals of understanding deep reinforcement learning for object tracking, teaching different types of learning, deep learning, reinforcement learning, policy, value function, approaches to reinforcement learning, action-driven object tracking, action-decision network, supervised learning, reinforcement learning, online adaptation, and self-comparison. The teaching method includes lectures and practical training exercises. The intended audience for this course is individuals interested in deep reinforcement learning for object tracking.

Syllabus

Intro
Different types of learning
Deep learning in a nutshell
Reinforcement learning in a nutshell
Policy
Value function A value function is a prediction of future reward
Approaches To Reinforcement Learning
Motivation
Action-driven object tracking
Problem definition (RL setting)
Action-decision network
Training: Supervised learning
Training: Reinforcement learning
Training: Online adaptation
Self comparison