万众瞩目！ 智能辅助助力e2e-MOT，全新Co-MOT横空出世！ 它突破了e2e-MOT的束缚，创新性地提出了基于合作竞争 ( COopetition)的标签分配策略，并引入影子概念，让检测query与追踪query互帮互助，大大提升了训练平衡性…

2023年5月31日 · To alleviate this problem, we present Co-MOT, a simple and effective method to facilitate e2e-MOT by a novel coopetition label assignment with a shadow concept. Specifically, we add tracked objects to the matching targets for detection queries when performing the label assignment for training the intermediate decoders.

2023年10月24日 · CO-MOT以高效的方式在多个数据集上实现了显著的性能提升，只需要 MOTRv2 38% 的 FLOPs 即可获得类似的性能，推理速度提高了 1.4 倍！ 现有的端到端多目标跟踪（e2e-MOT）方法尚未超越非端到端检测跟踪方法。

2023年9月21日 · To alleviate this problem, we present Co-MOT, a simple and effective method to facilitate e2e-MOT by a novel coopetition label assignment with a shadow concept. Specifically, we add tracked objects to the matching targets for detection queries when performing the label assignment for training the intermediate decoders.

To alleviate this problem, we present Co-MOT, a simple and effective method to facilitate e2e-MOT by a novel coopetition label assignment with a shadow concept. Specifically, we add tracked objects to the matching targets for detection queries when performing the label assignment for training the intermediate decoders.

aper proposes a method called CO-MOT to boost the performance of end-to-end Transformer-based MOT. We investigate the issues in the existing end-to-end MOT using Transformer and find that the label assignment can not

研究动机 多目标跟踪（Multiple Object Tracking，MOT）是计算机视觉尤其是视频理解中的一个重要任务，主要目标是定位目标并且在连续的视频帧中保持他们各自的身份信息（ID）。

智能辅助助力e2e-MOT，全新Co-MOT横空出世！ 它突破了e2e-MOT的束缚，创新性地提出了基于合作竞争 (COopetition)的标签分配策略，并引入影子概念，让检测query与追踪…

2023年10月22日 · 多目标跟踪 （Multiple Object Tracking，MOT）是 计算机视觉 尤其是视频理解中的一个重要任务，主要目标是定位目标并且在连续的视频帧中保持他们各自的身份信息（ID）。

2025年2月21日 · In this paper, we propose a GNN based 3D MOT method which effectively utilizes the collective motion consistency in traffic flow. Collective motion is modeled with a densely connected intra-flow graph within the collective group, allowing information to flow quickly.