Research map / Detection and segmentation / CNN features
Gold-YOLO: Efficient Object Detector via Gather-and-Distribute Mechanism
Chengcheng Wang, Wei He, Ying Nie, Jianyuan Guo, Chuanjian Liu, Yunhe Wang, Kai Han
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 |
|---|---|---|---|
| replace the YOLOv6-3.0 neck with Gold-YOLO's GD neck | Gold-YOLO-N · COCO 2017 val | AP: 37.0% -> 39.6% | Improved |
| add MIM pretraining | Gold-YOLO-S · COCO 2017 val | AP: 45.4% -> 45.5% | Improved |
| add MIM pretraining | Gold-YOLO-M · COCO 2017 val | AP: 49.8% -> 50.2% | Improved |
| add MIM pretraining | Gold-YOLO-L · COCO 2017 val | AP: 51.8% -> 52.3% | Improved |
| add High-GD to Low-GD (ablation) | Gold-YOLO-S · COCO val2017 | AP: 44.65% -> 45.57% | Improved |
| add LAF to Low-GD + High-GD (ablation) | Gold-YOLO-S · COCO val2017 | AP: 45.57% -> 46.11% | Improved |
| replace LAF concat with add (ablation) | Gold-YOLO-S · COCO val2017 | AP: 46.11% -> 45.43% | Hurt |
| increase model size with add-based LAF (ablation) | Gold-YOLO-S · COCO val2017 | AP: 45.43% -> 46.49% | Improved |
| remove LAF while increasing model size (ablation) | Gold-YOLO-S · COCO val2017 | AP: 46.49% -> 46.47% | No clear effect |
| increase MIM pretraining from 400 to 800 epochs (ablation) | Gold-YOLO-S · COCO 2017 val, no self-distillation | AP: 45.39% -> 45.52% | Improved |
| replace FPN with GD (ablation) | MaskRCNN-ResNet50 · COCO instance segmentation | Bbox mAP: 38.2% -> 40.7% | Improved |
| replace FPN with GD (ablation) | PointRend-ResNet50 · Cityscapes | mIoU: 76.47 -> 78.54 | Improved |
Closest papers by meaning
- YOLOv10: Real-Time End-to-End Object Detection (NeurIPS 2024 · similarity 0.90)
- Mamba YOLO: A Simple Baseline for Object Detection with State Space Model (AAAI 2025 · similarity 0.89)
- YOLO-World: Real-Time Open-Vocabulary Object Detection (CVPR 2024 · similarity 0.88)
- YOLO-RD: Introducing Relevant and Compact Explicit Knowledge to YOLO by Retriever-Dictionary (ICLR 2025 · similarity 0.88)
- YOLOv12: Attention-Centric Real-Time Object Detectors (NeurIPS 2025 · similarity 0.87)
- DETRs Beat YOLOs on Real-time Object Detection (CVPR 2024 · similarity 0.87)
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.
