CS598F: Advanced Topics in Computer Science: Deep Learning for Graphics and Vision

HEAD> CS598F: Advanced Topics in Computer Science: Deep Learning for Graphics and Vision

Princeton University
Computer Science Dept.

Computer Science 598F
Advanced Topics in Computer Science:
Deep Learning for Graphics and Vision

Spring 2017

General Information:

Time: Fridays 1:30-3:50PM
Place: CS 302
Instructor: Thomas Funkhouser
Piazza: web page

This course will survey recent research related to use of deep learning in computer graphics and computer vision. The course will operate as a graduate-level seminar with weekly readings, summarizations, and discussions of recent papers.

Schedule:

Date	Topics	Overview Presenter	Paper Presenter	Research Presenter	Required Readings
Feb 10	Introduction	Tom
Feb 17	Generative Adversarial Networks	Riley	Andras		[goodfellow2016nips] [shrivastava2016learning]
Feb 24	Learned Representations of 3D Shape	Ghassen and Jerry	Ghassen	Andreas Geiger (guest lecture)	[wu2016learning] [rezende2016unsupervised]
Mar 3	Self-Supervised Learning	Andy	Andy	Riley	[pathak2016context] [agrawa2015learning] [pinto2016supersizing]
Mar 10	Learning from RGB-D Panorama			Team	[armeni2017joint] [armeni20163d]
Mar 24	Spring Break
Mar 31	Deep Reinforcement Learning	Karan	Karan	Manolis	[mnih2013playing] [mnih2016asynchronous] [dosovitskiy2016learning]
Apr 7	Cross-Modal Learning with Text	Angel	Angel	Angel	[mao2016generation] [andreas2016learning] [wang2016learning]
Apr 14	Learning from Video	Kyle	Kyle		[karpathy2014large] [mathieu2015deep] [yeung2016end]
Apr 21	Learning for Robotic Manipulation	Karan	Karan and Andy	Andy	[lenz2015deep] [pinto2016curious] [pinto2017supervision]
Apr 21	Learning from Physics	Ghassen	Ghassen		[lerer2016learning] [agrawal2016learning]
May 5	Learning for Autonomous Driving	Alain, et al.	Andras		[janai2017computer]

Resources:

Bibliography:
- COS598F shared bibtex file (html)

Related courses:
- Representation Learning in Computer Vision, Silvio Savarese and Amir Zamir, Stanford University, Spring 2016.
- Deep Learning for Computer Graphics, Connelly Barnes, University of Virginia, Fall 2016
- Advanced Computer Vision, James Hays, Georgia Tech, Spring 2017
- Designing, Visualizing and Understanding Deep Neural Networks, John Canny, UC Berkeley, Fall 2016
- Computer vision seminar, University of Washington Vision Group, U Washington, Fall 2016
- Deep learning seminar, Philip Krahenbuhl, UT Austin, Fall 2016
- Visual Object and Activity Recognition, Trevor Darrell and Alyosha Efros, UC Berkeley, Fall 2016
- Visual Recognition, Kristen Grauman UT Austin, Fall 2016

Software Libraries:
- Caffe (C++ with Python bindings)
- Torch (Lua)
- TensorFlow (C++ with Python bindings)
- Keras, built on Theano (Python)