Learning to Recommend Frame for Interactive Video Object Segmentation in the Wild

The paper proposes a framework for interactive video object segmentation (VOS) in the wild, where users can choose some frames for annotations iteratively.

BibTex: