Research map / Detection and segmentation / CNN features
Does YOLO Really Need to See Every Training Image in Every Epoch?
Xingxing Xie, Jiahua Dong, Junwei Han, Gong 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 |
|---|---|---|---|
| use AFSS (adaptive image sampling) for YOLO training | YOLO11s · MS COCO 2017 | AP: 47.0 -> 47.2 | Improved |
| use AFSS for YOLO training speedup | YOLO11s · MS COCO 2017 | Training time (h): 43.9 -> 28.4 | Improved |
| use AFSS on oriented YOLO | YOLO11s-OBB · DOTA-v1.0 | mAP: 79.5 -> 79.6 | No clear effect |
| use AFSS instead of curriculum learning (ablation) | YOLO11s · MS COCO 2017 | AP: 43.7 -> 47.2 | Improved |
| use AFSS instead of data pruning (ablation) | YOLO11s · MS COCO 2017 | AP: 40.5 -> 47.2 | Improved |
| use AFSS instead of dataset distillation (ablation) | YOLO11s · MS COCO 2017 | AP: 35.6 -> 47.2 | Improved |
| add continuous review to LSM-only sampling (ablation) | AFSS · MS COCO 2017 | AP: 44.8 -> 45.5 | Improved |
| add state update to LSM+CR+STC (ablation) | AFSS · MS COCO 2017 | Speedup: 1.26x -> 1.54x | Improved |
| use min(Prec,Rec) sufficiency instead of loss-based (ablation) | AFSS · MS COCO 2017 | AP: 46.0 -> 47.2 | Improved |
| increase continuous review interval from 10 to 20 (ablation) | AFSS · MS COCO 2017 | AP: 47.2 -> 44.8 | Hurt |
| reduce continuous review interval from 10 to 5 (ablation) | AFSS · MS COCO 2017 | AP: 47.2 -> 47.2 | No clear effect |
| increase state update interval from 5 to 15 (ablation) | AFSS · MS COCO 2017 | AP: 47.2 -> 43.7 | Hurt |
Closest papers by meaning
- YOLOv10: Real-Time End-to-End Object Detection (NeurIPS 2024 · similarity 0.85)
- Mamba YOLO: A Simple Baseline for Object Detection with State Space Model (AAAI 2025 · similarity 0.84)
- YOLO-RD: Introducing Relevant and Compact Explicit Knowledge to YOLO by Retriever-Dictionary (ICLR 2025 · similarity 0.83)
- FBRT-YOLO: Faster and Better for Real-Time Aerial Image Detection (AAAI 2025 · similarity 0.83)
- YOLO-World: Real-Time Open-Vocabulary Object Detection (CVPR 2024 · similarity 0.83)
- Revisiting Generative Replay for Class Incremental Object Detection (CVPR 2025 · similarity 0.83)
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.
