Video instance segmentation (VIS) is a challenging vision task that aims to detect, segment, and track objects in videos. Conventional VIS methods rely on densely-annotated object masks which are expensive. We reduce the human annotations to only one point for each object in a video frame during training, and obtain high-quality mask predictions close to fully supervised models. Our proposed training method consists of a class-agnostic proposal generation module to provide rich negative samples and a spatio-temporal point-based matcher to match the object queries with the provided point annotations. Comprehensive experiments on three VIS benchmarks demonstrate competitive performance of the proposed framework, nearly matching fully supervised methods.
@article{arxiv.2404.01990,
title = {What is Point Supervision Worth in Video Instance Segmentation?},
author = {Shuaiyi Huang and De-An Huang and Zhiding Yu and Shiyi Lan and Subhashree Radhakrishnan and Jose M. Alvarez and Abhinav Shrivastava and Anima Anandkumar},
journal= {arXiv preprint arXiv:2404.01990},
year = {2024}
}