Research map / Detection and segmentation / CNN features
Visible and Clear: Finding Tiny Objects in Difference Map
Bing Cao, Haiyu Yao, Pengfei Zhu, Qinghua Hu
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 DGFE module on top of reconstruction head (ablation) | Cascade R-CNN · DroneSwarms | AP: 36.5 -> 38.3 | Improved |
| add reconstruction head alone (no difference map guidance) (ablation) | Cascade R-CNN · DroneSwarms | AP: 36.4 -> 36.5 | No clear effect |
| use element-wise attention instead of element-wise multiplication (ablation) | DGFE · DroneSwarms | AP: 36.2 -> 38.3 | Improved |
| use element-wise attention instead of concatenation (ablation) | DGFE · DroneSwarms | AP: 36.6 -> 38.3 | Improved |
| add learnable threshold filtration vs none (ablation) | DGFE · VisDrone2019 | AP: 27.0 -> 27.3 | Improved |
| make threshold learnable instead of fixed (ablation) | DGFE · VisDrone2019 | AP: 27.1 -> 27.3 | Improved |
| use high-frequency difference map instead of pixel difference map (ablation) | SR-TOD · DroneSwarms | AP: 38.3 -> 38.4 | No clear effect |
| add SR-TOD to Cascade R-CNN | Cascade R-CNN · DroneSwarms | AP: 36.4 -> 38.3 | Improved |
| add SR-TOD to DINO | DINO · DroneSwarms | AP: 35.4 -> 35.6 | Improved |
| add SR-TOD to RFLA | RFLA · VisDrone2019 | AP: 27.2 -> 27.8 | Improved |
| add SR-TOD to DetectoRS | DetectoRS · AI-TOD | AP: 14.6 -> 24.0 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 on most shared results.
- DetectoRS — DetectoRS: Detecting Objects With Recursive Feature Pyramid and Switchable Atrous Convolution (CVPR 2021) · better on 2 of 2 shared results
- Cascade R-CNN — Cascade R-CNN: Delving Into High Quality Object Detection (CVPR 2018) · better on 2 of 2 shared results
- Faster R-CNN — Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks (TPAMI 2016) · better on 2 of 2 shared results
- Deformable-DETR — Deformable DETR: Deformable Transformers for End-to-End Object Detection (ICLR 2021) · better on 1 of 1 shared result
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
- Oriented Tiny Object Detection: A Dataset, Benchmark, and Dynamic Unbiased Learning (TPAMI 2025 · similarity 0.86)
- Visible-Thermal Tiny Object Detection: A Benchmark Dataset and Baselines (TPAMI 2025 · similarity 0.86)
- Event-based Tiny Object Detection: A Benchmark Dataset and Baseline (ICCV 2025 · similarity 0.86)
- Learning 1-Bit Tiny Object Detector with Discriminative Feature Refinement (ICML 2024 · similarity 0.86)
- RemDet: Rethinking Efficient Model Design for UAV Object Detection (AAAI 2025 · similarity 0.85)
- SET: Spectral Enhancement for Tiny Object Detection (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.
