Adaptive ROI generation for video object segmentation using reinforcement learning

作者：

Highlights：

• We deploy the “actor critic” RL framework to train the agent for generating the adaptation areas.

• W e further design a novel multi branch tree based exploration method to fast select the best state action pairs to speed up the RL training.

• New state of the art result of mean re gion similarity is obtained for the DAVIS 2016 dataset, which is 87.1%.

摘要

•We deploy the “actor critic” RL framework to train the agent for generating the adaptation areas.•W e further design a novel multi branch tree based exploration method to fast select the best state action pairs to speed up the RL training.•New state of the art result of mean re gion similarity is obtained for the DAVIS 2016 dataset, which is 87.1%.

论文关键词：Model adaptation,Video object segmentation,Reinforcement learning,Training accelerate

论文评审过程：Received 5 October 2019, Revised 19 March 2020, Accepted 17 May 2020, Available online 24 May 2020, Version of Record 30 May 2020.

论文官网地址：https://doi.org/10.1016/j.patcog.2020.107465