Research map / Adversarial robustness and security / Attacks on text and graphs
PapMOT: Exploring Adversarial Patch Attack against Multiple Object Tracking
Jiahuan Long, Tingsong Jiang, Wen Yao, Shuai Jia, Weijia Zhang, Weien Zhou, Chao Ma, Xiaoqian Chen
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 bounding box restriction loss L_bbr (ablation) | PapMOT · MOT17 ByteTrack | IOR: 7.2 -> 79.5 | Improved |
| add average precision loss L_ap (ablation) | PapMOT · MOT17 FairMOT | TASR: 11.0 -> 76.8 | Improved |
| add L_bbr (ablation) | PapMOT · MOT17 ByteTrack | TASR: 61.3 -> 94.6 | Improved |
| add L_ap (ablation) | PapMOT · MOT17 ByteTrack | TASR: 21.9 -> 94.6 | Improved |
| use PapMOT instead of Trasw on MOT15 and MOT20 identity obfuscation | MOT attacks · MOT15 FairMOT | IOR: 66.2 (Trasw) vs 62.3 (PapMOT) | Hurt |
| use physical patch attack PapMOT instead of ObjPatch on BDD100K | ByteTrack · BDD100K | TASR: 20.5 -> 63.4 | Improved |
| use unoptimized patches (white, black, real image) instead of optimized patch | MOT physical attack · physical world | ID switch | No clear effect |
| increase patch size | PapMOT physical patch · physical world | attack strength | Improved |
| increase patch angle beyond ~15 degrees or distance beyond 4.1m | PapMOT physical patch · physical world | attack effectiveness: effective range 0.8m-4.1m distance | Hurt |
Closest papers by meaning
- Towards Universal Physical Attacks on Single Object Tracking (AAAI 2021 · similarity 0.89)
- Fooling Detection Alone is Not Enough: Adversarial Attack against Multiple Object Tracking (ICLR 2020 · similarity 0.89)
- APRICOT: A Dataset of Physical Adversarial Attacks on Object Detection (ECCV 2020 · similarity 0.88)
- Revisiting Adversarial Patch Defenses on Object Detectors: Unified Evaluation, Large-Scale Dataset, and New Insights (ICCV 2025 · similarity 0.88)
- Revisiting Adversarial Patches for Designing Camera-Agnostic Attacks against Person Detection (NeurIPS 2024 · similarity 0.86)
- PAD: Patch-Agnostic Defense against Adversarial Patch Attacks (CVPR 2024 · similarity 0.85)
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.
