Research map / Detection and segmentation / Tracking
The Unmanned Aerial Vehicle Benchmark: Object Detection and Tracking
Dawei Du, Yuankai Qi, Hongyang Yu, Yifan Yang, Kaiwen Duan, Guorong Li, Weigang Zhang, Qingming Huang, Qi Tian
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 |
|---|---|---|---|
| evaluate R-FCN on UAV imagery instead of KITTI hard set | R-FCN · UAVDT-DET vs KITTI-D hard | AP: 70.06 -> 34.35 | Hurt |
| evaluate MDNet on UAVDT instead of OTB100 | MDNet · UAVDT-SOT vs OTB100 | success score: 69.4 -> 46.4 | Hurt |
| evaluate KCF on UAVDT instead of UAV123 | KCF · UAVDT-SOT vs UAV123 | success score: 33.1 -> 29.0 | Hurt |
| use region-based detector R-FCN rather than region-free RON | object detectors · UAVDT-DET | AP | Improved |
| add RON reverse connection fusing later layers to SSD | SSD · UAVDT-DET | AP | Hurt |
| tune anchor sizes of author-provided detector settings for small objects | DET methods · UAVDT-DET | accuracy: 10% accuracy rise | Improved |
| replace Faster-RCNN detections with R-FCN detections for MDP | MDP · UAVDT-MOT | MOTA: 43.0 -> 28.9 | Hurt |
| add appearance features (DSORT) to SORT with Faster-RCNN detections | SORT · UAVDT-MOT (Faster-RCNN detections) | MOTA: 39.0 -> 40.7 | Improved |
| use daylight instead of night or fog sequences for MOT | MOT methods · UAVDT-MOT attributes | MOTA | Improved |
Closest papers by meaning
- UAV3D: A Large-scale 3D Perception Benchmark for Unmanned Aerial Vehicles (NeurIPS 2024 · similarity 0.93)
- The Unmanned Aerial Vehicle Benchmark: Object Detection, Tracking and Baseline (IJCV 2019 · similarity 0.93)
- RemDet: Rethinking Efficient Model Design for UAV Object Detection (AAAI 2025 · similarity 0.87)
- M$^2$E-UAV: A Benchmark and Analysis for Onboard Motion-on-Motion Event-Based Tiny UAV Detection (NeurIPS 2026 · similarity 0.87)
- MUST: The First Dataset and Unified Framework for Multispectral UAV Single Object Tracking (CVPR 2025 · similarity 0.86)
- Multiview Aerial Visual RECognition (MAVREC): Can Multi-view Improve Aerial Visual Perception? (CVPR 2024 · 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.
