DLCV - VISION & LEARNING LAB

Introduction

Computer Vision has become ubiquitous in our society, with applications in image/video search and understanding, apps, mapping, medicine, drones, and self-driving cars. Core to many of these applications are visual recognition tasks such as image classification, segmentation, localization and detection. Recent developments in neural network (a.k.a. deep learning) approaches have greatly advanced the performance of these state-of-the-art visual recognition systems. This course is a deep dive into details of the deep learning architectures with a focus on learning end-to-end models for solving these tasks.

Goals

This course will expose students to cutting-edge research — starting from a refresher in basics of machine learning, computer vision, neural networks, to recent developments. The emphasis will be on student-led paper presentations and discussions. Each topic will begin with instructor lectures to present context and background material.

Syllabus

Week	Date	Topic	Course Materials	Remarks
0	02/20	Course Logistics + Intro to Computer Vision	[Slides]
1	02/27	Machine Learning 101	[Slides]
2	03/06	Image Representation: From Recognition to Tracking	[Slides]	HW #1 out
3	03/13	Intro to Neural Networks + CNN (I)	[Slides]
4	03/20	Intro to Neural Networks + CNN (II) Tutorial on Python, PyTorch, Git, etc.	[Slides]	HW #1 due
5	03/27	Detection & Segmentation	[Slides]	HW #2 out
6	04/03	Spring Break
7	04/10	Segmentation, Visualization, and Adversarial Attacks of NNs	[Slides]	[]
8	04/17	Generative Adversarial Networks (Guest Lecture by Prof. Wei-Chen Chiu)	[Slides]	HW #2 due HW #3 out
9	04/24	Transfer Learning and Representation Disentanglement (I)	[Slides]
10	05/01	Transfer Learning and Representation Disentanglement (II)	[Slides]	Team Up for Final Project
11	05/08	Deep Learning for Video Frame Interpolation & Object Co-Segmentation (Guest Lecture by Dr. Yen-Yu Lin)	[Slides]	HW #3 due
12	05/15	Recurrent NNs and Seq-to-Seq Models (I)	[Slides]	HW #4 out
13	05/22	Recurrent NNs and Seq-to-Seq Models (II); Invited talk by Dr. Taiyi Huang (ASUS AICS)	[Slides]
14	05/29	Learning Beyond Images (2D/3D, depth, etc.)	[Slides]
15	06/05	Audio-Visual Event Localization and Synthesis; Invited talk (1) by Dr. Wei-Sheng Lai, UC Merced) Invited talk (2) by Dr. Shang-Hong Lai (Principal Researcher, Microsoft)	[Slides]	HW #4 due
16	06/12	Few-Shot Learning	[Slides]	[]