Research map / Detection and segmentation / Open-set and few-shot detection
Visual Prototype Conditioned Focal Region Generation for UAV-Based Object Detection
Wenhao Li, Zimeng Wu, Yu Wu, Zehua Fu, Jiaxin 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 UAVGen synthesized data to real training data | GFL-ResNet50 · VisDrone | mAP: 24.5 -> 25.9 | Improved |
| add UAVGen synthesized data to real training data | GFL-ResNet50 · UAVDT | mAP: 14.5 -> 16.6 | Improved |
| add CopyPaste augmentation | GFL · VisDrone | mAP: 24.5 -> 23.6 | Hurt |
| add GLIGEN/Geodiffusion/AeroGen synthesized data | RemDet-X · VisDrone | mAP: 29.8 -> 28.8 (AeroGen) | Hurt |
| add UAVGen synthesized data | RemDet-X · VisDrone | mAP: 29.8 -> 30.2 | Improved |
| add generated images without visual prototype or layout embedding (ablation) | GFL · VisDrone | mAP: 24.5 -> 23.8 | Hurt |
| add visual prototypes and layout embeddings to the diffusion model (ablation) | VPC-DM · VisDrone | mAP: 23.8 -> 25.2 | Improved |
| add focal region enhanced data pipeline (focal region and label refinement) (ablation) | UAVGen on VPC-DM · VisDrone | mAP: 25.2 -> 25.9 | Improved |
| lower focal region resolution from 1024 to 256 (ablation) | UAVGen · VisDrone | mAP: 25.3 -> 25.9 | Improved |
| add visual prototype conditioning to layout-to-image generation | UAVGen vs AeroGen · VisDrone | FID: 48.04 -> 34.34 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- AeroGen — AeroGen: Enhancing Remote Sensing Object Detection with Diffusion-Driven Data Generation (CVPR 2025) · better on 7 of 7 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- RemDet: Rethinking Efficient Model Design for UAV Object Detection (AAAI 2025 · similarity 0.86)
- The Unmanned Aerial Vehicle Benchmark: Object Detection and Tracking (ECCV 2018 · similarity 0.86)
- UAV3D: A Large-scale 3D Perception Benchmark for Unmanned Aerial Vehicles (NeurIPS 2024 · similarity 0.85)
- ODGEN: Domain-specific Object Detection Data Generation with Diffusion Models (NeurIPS 2024 · similarity 0.84)
- Learning Adaptive and View-Invariant Vision Transformer for Real-Time UAV Tracking (ICML 2024 · similarity 0.84)
- Similarity-Guided Layer-Adaptive Vision Transformer for UAV Tracking (CVPR 2025 · similarity 0.84)
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.
