TMVOS: Triplet Matching for Efficient Video Object Segmentation

作者：

Highlights：

• A new triplet matching strategy is devised to generate accurate matching maps.

• A dual decoder is implemented to accomplish the final segmentation.

• The proposed model well fits the complex background and relatively simple targets.

• Our method achieves promising performance on effectiveness and efficiency than SOTA.

摘要

•A new triplet matching strategy is devised to generate accurate matching maps.•A dual decoder is implemented to accomplish the final segmentation.•The proposed model well fits the complex background and relatively simple targets.•Our method achieves promising performance on effectiveness and efficiency than SOTA.

论文关键词：Video object segmentation,Embedding learning,Triplet matching

论文评审过程：Received 24 July 2021, Revised 31 March 2022, Accepted 8 June 2022, Available online 15 June 2022, Version of Record 28 June 2022.

论文官网地址：https://doi.org/10.1016/j.image.2022.116779