Tag
TCoVP 引入时空绑定的视频推理链(Temporal Chain of Video Primitives),将实体引用与带时间戳的空间证据关联,用于视频推理分割任务,并通过 SFT 与链对齐强化学习数据集在多个基准上取得提升。
LVMT (Long-term Video Mask Transformer) introduces a GRU-based temporal propagation module and a Truncated Query Propagation (TQP) training strategy to handle long-term occlusions in online video segmentation, setting new state-of-the-art results across six benchmarks while running 10x faster than prior methods.
360 AI Research Institute's MoSA (Motion-Grounded Segment Anything), accepted at ECCV 2026, trains the Segment Anything Model to segment objects using motion cues from unlabeled video, generating millions of pseudo-labels without manual annotation.
This paper introduces SAM-MT, an extension of SAM2 for real-time interactive multi-target video segmentation, achieving high FPS independent of target count.