Prof. Lo and His Team Won the Best Demonstration Runner Up Award at ACM SIGMOD 2022

Prof. Eric Lo and his team members (Ziliang Lai, Chris Liu, Chenxia Han, Pengfei Zhang and Ben Kao) won the best demonstration runner up award at ACM SIGMOD 2022 with “Everest: A Top-K Deep Video Analytics System“.

The SIGMOD demonstration program is for sharing cutting-edge, data management system prototypes with the greater SIGMOD community. Demonstrations may consist of software, hardware, or both. The emphasis of the SIGMOD demonstration program is on visionary, next-generation systems requiring significant research and development effort.

Everest returns the highest-scored frames (or short clips) for your uploaded videos. Scores of every frame/clip are accessed by a deep model. Currently, Everest has three built-in deep models, which can extract the funniest moment of your home video, the crowdiest moment of a traffic footage, and the most exciting moment of a soccer match. Everest supports UDF so that you can add any new model at ease. Everest promises to return you the answer real quick, using minimal GPU resources with high accuracies.

 

Everest: A Top-K Deep Video Analytics System
Everest: A Top-K Deep Video Analytics System