How can we conduct multimodal multi-objective tracking and how is it evaluated?

open-mmlab / mmtracking

OpenMMLab Video Perception Toolbox. It supports Video Object Detection (VID), Multiple Object Tracking (MOT), Single Object Tracking (SOT), Video Instance Segmentation (VIS) with a unified framework.

https://mmtracking.readthedocs.io/en/latest/

Apache License 2.0

3.58k stars 597 forks source link

How can we conduct multimodal multi-objective tracking and how is it evaluated? #941

Open wensir66666 opened 1 month ago

wensir66666 commented 1 month ago

How can we conduct multimodal multi-objective tracking and how is it evaluated?