image understanding at Google research

Developed/applied machine learning/deep learning, computer vision and image processing techniques in solving real-world problems at Facebook scale.

Implemented and optimized an adaptive video denoising filter that reduces noise while preserving

edges and salient regions of interests.

Conducted research and development on future multimedia standards (FMS) or next generation video coding (NGVC) project and 3D video project.

Tracked the video standardization activities such as MPEG, ITU-T (VCEG) and AVS standards, especially on high efficient video codec standard (HVC) and 3D video.

Helped universities transform their research achievements of the FMS or NGVC project and 3D video project into applications.

Worked on the Video Post Processing (VPP) project, with an emphasis on Deinterlacing and Frame Rate Up Conversion (FRUC).