About Me
Hi! My name is Anton, I currently work as a Software Engineer at Google, and also pursue academic research. At Google, I work on a machine learning system for modeling and predicting user behavior. A lot of my work is concentrated on improving accuracy, quality and efficiency of inference and training systems.
My research interests lie in the fields of Computer Vision and Decision Making. In particular, I am most interested in using generative approaches to build novel human-centered systems, as well as developing agents that laverage uncertainty in novel environments!
I have worked on research projects in projective geometry with Dr. Annalisa Crannel (Franklin and Marshall College) and artificial intelligence with Dr. Erin Talvitie (Harvery Mudd College). I also worked on research at the GRASP Lab under supervison of Dr. Kostas Daniilidis (University of Pennsylvania), and collaborate with Dr. Bernadette Bucher (University of Michigan). I also collaborate with the research group of Dr. Junchen Jiang at the University of Chicago.
The main focus of my research has been the use of uncertainty and generative models in two fields: live video codecs and autonomus agent navigation in novel environments.
News
Computer Science Department at the University of Chicago has recently released an article about our most recent project "GRACE": UChicago Computer Scientists Bring in Generative Neural Networks to Stop Real-Time Video From Lagging
Select Publications
For the full list of my publications and pre-prints see my Google Scholar- Yihua Cheng, Ziyi Zhang, Hanchen Li, Anton Arapin, Yue Zhang, Qizheng Zhang, Yuhan Liu, Kuntai Du, Xu Zhang, Francis Y. Yan, Amrita Mazumdar, Nick Feamster, Junchen Jiang. "GRACE: Loss-Resilient Real-Time Video through Neural Codecs", NSDI 2024.
- Yihua Cheng, Anton Arapin, Ziyi Zhang, Qizheng Zhang, Hanchen Li, Nick Feamster, Junchen Jiang. "Grace++: Loss-Resilient Real-Time Video Communication under High Network Latency" CoRR 2023.
- Kuntai Du, Qizheng Zhang, Anton Arapin, Haodong Wang, Zhengxu Xia, Junchen Jiang. "Accmpeg: Optimizing video encoding for accurate video analytics", MLSys 2022.
- Georgios Georgakis, Bernadette Bucher, Anton Arapin, Karl Schmeckpeper, Nikolai Matni, Kostas Daniilidis. "Uncertainty-driven planner for exploration and navigation", ICRA 2022.
- Bernadette Bucher, Anton Arapin, Ramanan Sekar, Marc Badger, Feifei Duan, Oleh Rybkin, Kostas Daniilidis. "Perception-driven curiosity with bayesian surprise", RSS Workshop 2019.
Other Interests
I like studying other cultures and languages, I speak English, Ukrainian, Russian, Japanese and just started learning Spanish!I am also fascinated with art history and folk arts, examples of topics I have been reading about for the past few yers: customs and cultures of indigenous Ainu people from Hokkaido, Japanese woodblock prints (I visit every new exhibit at the Art Institute of Chicago and own a few myself!), and ancient Chinese bronze, especially the mysterious Jinsha culture.
I love physical exercises, and practice a few matrial arts: Taekwondo (hopefully will get my black belt this year!), Boxing, Brazilian Jiu Jitsu (Just started a year ago and loving it so far).
Contact Me
If you wish to get in touch for collaboration or inquiries, please reach out:
- Email: anton dot arapin at gmail dot com
- LinkedIn: LinkedIn
- Google Scholar: Google Scholar