Costa Huang's Website

Posted on Sat, Feb 15, 2020

πŸ‘‹ Hi, I'm Costa, a Computer Science Ph.D. student at Drexel University studying Game Artificial Intelligence. I enjoy doing research and conducting super cool experiments!

🐦 Twitter πŸ‘¨β€πŸ’Ό Linkedin πŸ™ Github πŸ“œ Resume

I use Deep Reinforcement Learning (DRL) to train bots to play games autonomously!

Our trained agent learned to play breakout!!

My bot learns to play volleyball through self-play

My more recent work focuses on scaling DRL to Real-time Strategy (RTS) games, in collaboration with my advisor Santiago OntaΓ±Γ³n.

Let's safely land our shuttle to the lunar surface. Gotta be steady and precise!

Have some fun with a car racing game!

A simulated RTS game. The white, green block and circles are bases, resources, and workers. The purpose of the game is to gather resources, build an army and destroy enemy forces.

You can easily use my DRL library CleanRL to train the agents to play all of the games above.

Feel free to get in touch with me at [email protected] πŸŽ‰.

πŸ”§ My Projects

Feel free to click on the links for details

πŸ“– My Publications

Huang, S., OntaΓ±Γ³n, S., "Action Guidance: Getting the Best of Sparse Rewards and Shaped Rewards for Real-time Strategy Games", AIIDE Workshop on Artificial Intelligence for Strategy Games, October 2020

Huang, S., OntaΓ±Γ³n, S., "A Closer Look at Invalid Action Masking in Policy Gradient Algorithms", Preprint.

Huang, S., OntaΓ±Γ³n, S., β€œComparing Observation and Action Representations for Reinforcement Learning in Β΅RTS”, AIIDE Workshop on Artificial Intelligence for Strategy Games, October 2019

Huang, S., Grethlein, D., β€œGenerating Interpretable Class Model Visualizations for CNNs with varying Dilation Factors”, preprint, June 2019

Huang, S., Healy, C., β€œStreetTraffic: a Library for Traffic Flow Data Collection and Analysis”, ACMSE 2018 Conference, March 2018

πŸ“œ My Resume