Action Recognition——Deep Domain Adaptation in Action Space——BMVC2018

Abstract

The problem of Domain Shift in action videos.

Introduction

实际应用例子surveillance cameras are everywhere, be it city streets, market place, buildings or airports. A massive amount of video data that needs to be processed for autonomous understanding of events and activities occurring in the scene.
Model the action video as a sequence of points on the subspace

Conclusion

formulate the problem of domain adaptation for human action recognition; future study the concept of continuous domain adaptation on the streaming action videos

Key points: the paper写作一般；