Li Ding | 丁立


I'm currently at MIT, working on deep learning for perception and control of autonomous vehicles. Our research focuses on Human-Centered Artificial Intelligence (HCAI) that leverages human knowledge to enhance machine intelligence.

I TAed MIT 6.S094: Deep Learning for Self-driving Cars and MIT 6.S099: Artificial General Intelligence in Jan. & Feb. 2018. If you have questions regarding the course, please email us through deepcars [at] mit.edu / agi [at] mit.edu, respectively.

Prior to joining MIT, I worked on deep learning for human action recognition at University of Rochester (Dept. of Computer Science), after getting a Master degree in Data Science.

I'm from Shanghai, China. On a side of fun, I'm a casual Kaggler interested in playing with various kinds of data. I like photography, electro-funk, all kinds of cuisine, and at the moment, walking and traveling around with Pokémon Go.

News: One paper on weakly supervised action recognition has been accepted to CVPR '18, arXiv and code available.

Github | LinkedIn
liding [at] mit.edu


Research
Ongoing

Driving Scene Perception

Adversarial Attack

Neural Synthesis

Dynamic Scene and Optical Flow


Past

Weakly Supervised Action Localization

Edge Cases in Image Recognition

Fine-grained Image Classification


Publication

Weakly-Supervised Action Segmentation with Iterative Soft Boundary Assignment
Li Ding, Chenliang Xu
[CVPR '18] [arXiv: 1803.10699] [Code: Github]

MIT Autonomous Vehicle Technology Study: Large-Scale Deep Learning Based Analysis of Driver Behavior and Interaction with Automation
Lex Fridman, Daniel E. Brown, Michael Glazer, William Angell, Spencer Dodd, Benedikt Jenik, Jack Terwilliger, Julia Kindelsberger, Li Ding, Sean Seaman, Hillary Abraham, Alea Mehler, Andrew Sipperley, Anthony Pettinato, Bobbie Seppelt, Linda Angell, Bruce Mehler, Bryan Reimer
[arXiv: 1711.06976]

TricorNet: A Hybrid Temporal Convolutional and Recurrent Network for Video Action Segmentation
Li Ding, Chenliang Xu
[arXiv: 1705.07818]


Service
TA


Reviewer

IEEE Transactions on Circuits and Systems for Video Technology (2018)
IEEE Access (2018)


Kaggle

Level: Competitions Expert (highest rank: 1169 | current rank)

Statoil/C-CORE Iceberg Classifier Challenge
(Satellite Image Classification)

· 2018 · Top 6%

Data Science Bowl 2017
(Lung Cancer Detection)

· 2017 · Top 6%


Thanks for visiting!