Research map / Detection and segmentation / CNN features
YOLO-ULM: Ultra-Lightweight Models for Real-Time Object Detection
Shasha Han, Chong Li, Xinning Wang, Xuebo Li
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 D3C2f feature aggregation (ablation) | YOLO-ULM-N · COCO val | AP50:95: 40.7 -> 40.9 | Improved |
| add D3C2f (ablation) | YOLO-ULM-M · COCO val | AP50:95: 52.6 -> 52.6 | No clear effect |
| add RepDown downsampling (ablation) | YOLO-ULM-N · COCO val | AP50:95 / params: 40.9 -> 41.0; params 2.6M -> 2.1M | Improved |
| add RepDown downsampling (ablation) | YOLO-ULM-M · COCO val | AP50:95 / params: 52.6 -> 52.5; params 20.2M -> 16.0M | Mixed |
| add FoCIoU loss (ablation) | YOLO-ULM-N · COCO val | AP50:95: 41.0 -> 41.2 | Improved |
| add FoCIoU loss (ablation) | YOLO-ULM-M · COCO val | AP50:95: 52.5 -> 52.8 | Improved |
| use FoCIoU instead of Focaler-IoU (ablation) | YOLO-ULM-N · COCO val | AP50:95: 40.9 -> 41.2 | Improved |
| use RepDown instead of ADown (ablation) | YOLO-ULM-N · COCO val | AP50:95: 40.7 -> 41.2 | Improved |
| use 7x7 kernel in RepDown instead of 5x5 or 9x9 (ablation) | YOLO-ULM-N · COCO val | AP50:95: 41.3 / 41.2 -> 41.6 | Improved |
| use padding in RepDown with 600 epochs (ablation) | YOLO-ULM-N · COCO val | AP50:95: 41.6 -> 41.2 | Hurt |
| use padding in RepDown with 600 epochs (ablation) | YOLO-ULM-S · COCO val | AP50:95: 47.8 -> 48.1 | Improved |
| use grouped convolution in downsampling (ablation) | YOLO-ULM-L/X · COCO val | params / AP50:95: L 54.1 -> 54.1; X 55.7 -> 55.6 | Improved |
Closest papers by meaning
- YOLOv10: Real-Time End-to-End Object Detection (NeurIPS 2024 · similarity 0.92)
- Mamba YOLO: A Simple Baseline for Object Detection with State Space Model (AAAI 2025 · similarity 0.90)
- YOLO-RD: Introducing Relevant and Compact Explicit Knowledge to YOLO by Retriever-Dictionary (ICLR 2025 · similarity 0.90)
- YOLO-World: Real-Time Open-Vocabulary Object Detection (CVPR 2024 · 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 0.86)
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.
