The 1st-place Solution for ECCV 2022 Multiple People Tracking in Group Dance Challenge
TL
TL;DR
This paper presents the 1st place solution to the Group Dance Multiple People Tracking Challenge, surpassing the second-place solution by +6.8% HOTA and using the YOLOX detection proposals for the anchor initialization of detect queries.
Abstract
We present our 1st place solution to the Group Dance Multiple People Tracking Challenge. Based on MOTR: End-to-End Multiple-Object Tracking with Transformer [13], we explore: 1) detect queries as anchors, 2) tracking as query denoising, 3) joint training on pseudo video clips generated from CrowdHuman dataset [10], and 4) using the YOLOX [4] detection proposals for the anchor initialization of detect queries. Our method achieves 73.4% HOTA on the DanceTrack test set, surpassing the second-place solution by +6.8% HOTA.
References15 items
1
Anchor DETR: Query Design for Transformer-Based Detector
2
Observation-Centric SORT: Rethinking SORT for Robust Multi-Object Tracking
3
DN-DETR: Accelerate DETR Training by Introducing Query DeNoising
4
DAB-DETR: Dynamic Anchor Boxes are Better Queries for DETR
5