AI RESEARCH
Bootstrapping Video Semantic Segmentation Model via Distillation-assisted Test-Time Adaptation
arXiv CS.CV
•
ArXi:2604.10950v1 Announce Type: new Fully supervised Video Semantic Segmentation (VSS) relies heavily on densely annotated video data, limiting practical applicability. Alternatively, applying pre-trained Image Semantic Segmentation (ISS) models frame-by-frame avoids annotation costs but ignores crucial temporal coherence. Recent foundation models such as SAM2 enable high-quality mask propagation yet remain impractical for direct VSS due to limited semantic understanding and computational overhead.