Research map / Detection and segmentation / Tracking
CoopTrack: Exploring End-to-End Learning for Efficient Cooperative Sequential Perception
Jiaru Zhong, Jiahao Wang, Jiahui Xu, Xiaofan Li, Zaiqing Nie, Haibao Yu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add fusion-after-decoding pipeline (ablation) | CoopTrack · V2X-Seq (10Hz) ablation | mAP: 0.310 -> 0.337 | Improved |
| add multi-dimensional feature extraction (MDFE) (ablation) | CoopTrack · V2X-Seq (10Hz) ablation | AMOTA: 0.277 -> 0.283 | Improved |
| add cross-agent alignment (CAA) (ablation) | CoopTrack · V2X-Seq (10Hz) ablation | AMOTA: 0.277 -> 0.304 | Improved |
| add graph-based association (GBA) (ablation) | CoopTrack · V2X-Seq (10Hz) ablation | AMOTA: 0.277 -> 0.295 | Improved |
| add GBA on top of MDFE and CAA (ablation) | CoopTrack · V2X-Seq (10Hz) ablation | AMOTA: 0.332 -> 0.346 | Improved |
| increase historical frames from 0 to 4 (ablation) | CoopTrack · V2X-Seq (10Hz) | AMOTA: 0.100 -> 0.346 | Improved |
| increase historical frames from 4 to 5 (ablation) | CoopTrack · V2X-Seq (10Hz) | AMOTA: 0.346 -> 0.324 | No clear effect |
| use end-to-end CoopTrack instead of UniV2X | CoopTrack · V2X-Seq (2Hz) | AMOTA: 0.239 -> 0.328 | Improved |
| use CoopTrack instead of V2X-ViT | CoopTrack · V2X-Seq (2Hz) | mAP: 0.268 -> 0.390 | Improved |
| use CoopTrack instead of Early Fusion | CoopTrack · Griffin-25m | AMOTA: 0.670 -> 0.488 | Hurt |
| replace ResNet101 with ConvNeXt-S backbone | CoopTrack · V2X-Seq (2Hz) | AMOTA: 0.328 -> 0.429 | Improved |
| add feature flow prediction for latency compensation | CoopTrack · V2X-Seq 500ms latency | mAP/AMOTA degradation: 15.3% -> 3.8% mAP drop; 31.8% -> 6.8% AMOTA drop | Improved |
Closest papers by meaning
- CoSDH: Communication-Efficient Collaborative Perception via Supply-Demand Awareness and Intermediate-Late Hybridization (CVPR 2025 · similarity 0.87)
- End-to-End Autonomous Driving Through V2X Cooperation (AAAI 2025 · similarity 0.86)
- Privacy-Preserving V2X Collaborative Perception Integrating Unknown Collaborators (AAAI 2025 · similarity 0.86)
- BEVSync: Asynchronous Data Alignment for Camera-based Vehicle-Infrastructure Cooperative Perception Under Uncertain Delays (AAAI 2025 · similarity 0.86)
- SparseAlign: a Fully Sparse Framework for Cooperative Object Detection (CVPR 2025 · similarity 0.86)
- STAMP: Scalable Task- And Model-agnostic Collaborative Perception (ICLR 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
