Columbia University E6894, Spring 2017 (7:00-9:30pm, Wednesday, 627 Seeley W. Mudd Building)

Deep Learning for Computer Vision, Speech, and Language

Tentative Course Schedule

Week Topic (first half) Topic (second half) Note
1 (1/18) Liangliang
Course overview
Xiaodong, Liangliang, Deep learming demos:
First homework:
please email your solution to Yizhou Wang
(yw2875_at_columbia.edu)
2 (1/25) Xiaodong
Basic math for neural networks
Kapil
Optimization techniques for neural networks
Liangliang
What we can do with deep learning toolkits
First homework deadline: Noon of Jan 25.
3 (2/1) Chad
Tutorial on Keras
In-class programming Extended homework deadline: Noon of Feb 1
(For those who joined late from the waiting list)
Registering student presentation slots at Google Group
4 (2/8) Oscar Chang
The best performer of in class programming context
Xiaodong
Deep Learning for Automatic Speech Recognition – Part I
Deadline of registering student presentation: Feb 9
Finalize your registration: Google spreadsheet
5 (2/15) Xiaodong
Deep Learning for Automatic Speech Recognition – Part II
Emily Hua and Kaili Chen: Convolutional Neural Networks for Speech Recognition
Liangliang: How to find an innovative project idea
6 (2/22) Markus Nussbaum-Thom
End-to-end speech recognition
student presentations:
Zixiaofan Yang and Xing Lan: End2End speech recognition with RNN
Wanting Wang: Neural Speech Recognizer
Oscar Chang and Siddharth Varia: Deep Speech2
Submitting final project team information here
7 (3/1) Kapil
Language Representation and Modeling
student presentations:
Elsbeth Turcan and Fei-Tzin Lee: Varational Autoencoders Write Poetry
Jon Koss and Abhishek Jindal: Fast Text
8 (3/8) Kapil
Sequence-to-Sequence Architectures
student presentations:
Anshul Sacheti and YangLu Piao, LSTM: A Search Space Odyssey
Sun Mao and Yu Zheng, Translation in Linear Time
9 (3/15) No class (spring break) No class (spring break)
10 (3/22) Kapil
Applications of natural language processing
student presentations:
Akshay Khatri
Apoorv Kulshreshtha and Samarth Tripathi: LSTM Networks for Machine Reading
Jose Ruiz and Pablo Vicente: Dueling network architectures for deep reinforcement learning
11 (3/29) Liangliang
Image recognition misc: model adaption, hyperface, and mask r-cnn
student presentations:
Mingyang Zheng and Lingyu Zhang: SqueezeNet
Wenxin Chen and Wen Zhang: Ignacio Aranguren and Rahul Rana: Face Net
Final project checkpoint
12 (4/5) Nikolai Yakovenko
Poker AI
student presentations: Chen-Yu Yen and Yu Chun Chien Deep Q-Network
Ruixuan Zhng and Xiang Hua, DeepStack for No-Limit Poker
Richard Godden and Yogesh Garg, Generative Adversarial Nets
13 (4/12) Shiyu Chang: WaveNet Student presentation:
Anthony Alvarez and GwonJae Cho: Google’s Neural Machine Translation System
Manu Gandham and Harish Shanker: Skip Thought Vectors
Hassan Akbari and Himani Arora: Visual Question Answering
14 (4/19) No class Preparing for the final project
15 (4/26) final project presentation at 750 CEPSR
(5/7) final project report due
Liangliang Cao Updated 04/11/2017