United States
Build intelligent machines that can learn a wide range of tasks as efficiently as humans. Natural language processing, computer vision, reinforcement learning, and more.
Designed, developed and deployed recommendation system capable of handling billions of objects and billion+ users.
Developed convolutional neural networks for a variety of visual understanding tasks, including human detection, facial key points detection, facial attribute recognition, human action classification, object tracking, etc. The models were actually deployed in many real systems, including the video surveillance at San Francisco Airport and some other US airports.