Research map / Detection and segmentation / CNN features
DQ-DETR: DETR with Dynamic Query for Tiny Object Detection
Yi-Xin Huang, Hou-I Liu, Hong-Han Shuai, Wen-Huang Cheng
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 categorical counting, dynamic query selection and feature enhancement to DINO-DETR (ablation) | DINO-DETR · AI-TOD-V2 test | AP: 25.9 -> 30.2 | Improved |
| add counting-guided feature enhancement to counting plus dynamic queries (ablation) | DQ-DETR · AI-TOD-V2 test | AP: 28.1 -> 30.2 | Improved |
| add dynamic query selection to counting plus feature enhancement (ablation) | DQ-DETR · AI-TOD-V2 test | AP: 29.1 -> 30.2 | Improved |
| add counting and dynamic queries without feature enhancement (ablation) | DINO-DETR · AI-TOD-V2 test | AP_vt: 12.7 -> 12.3 | Hurt |
| regress instance count instead of classifying into four levels (ablation) | categorical counting module · AI-TOD-V2 test | AP: 30.2 -> 14.9 | Hurt |
| split counts into five classes instead of four (ablation) | categorical counting module · AI-TOD-V2 test | AP: 30.2 -> 28.8 | Hurt |
| use dynamic number of queries instead of fixed 900 for images with N <= 10 (ablation) | DQ-DETR vs DINO-DETR · AI-TOD-V2, N <= 10 | AP: 22.5 -> 26.1 | Improved |
| use 1500 queries for very dense images instead of fixed 900 (ablation) | DQ-DETR vs DINO-DETR · AI-TOD-V2, N > 500 | AP: 13.5 -> 14.7 | Improved |
| apply DQ-DETR to general detection (COCO) instead of DINO-DETR | DINO-DETR · COCO val | AP: 51.3 -> 50.2 | Hurt |
| apply DQ-DETR instead of DINO-DETR on VisDrone | DINO-DETR · VisDrone val | AP: 35.8 -> 37.0 | 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.
- 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
- MS-DETR: Efficient DETR Training with Mixed Supervision (CVPR 2024 · similarity 0.88)
- Q-DETR: An Efficient Low-Bit Quantized Detection Transformer (CVPR 2023 · similarity 0.87)
- DETRs Beat YOLOs on Real-time Object Detection (CVPR 2024 · similarity 0.86)
- V-DETR: DETR with Vertex Relative Position Encoding for 3D Object Detection (ICLR 2024 · similarity 0.86)
- MI-DETR: An Object Detection Model with Multi-time Inquiries Mechanism (CVPR 2025 · similarity 0.86)
- AQ-DETR: Low-Bit Quantized Detection Transformer with Auxiliary Queries (AAAI 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.
