Commit ·
3a80764
1
Parent(s): 18a6586
update README.md
Browse files
README.md
CHANGED
|
@@ -16,6 +16,16 @@ inference: false
|
|
| 16 |
|
| 17 |
This repository contains five trained PyTorch D-FINE models proposed in [D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement](https://arxiv.org/abs/2410.13842) by Yansong Peng, Hebei Li, Peixi Wu, Yueyi Zhang, Xiaoyan Sun, Feng Wu.
|
| 18 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
|
| 20 |
## Citation
|
| 21 |
If you use `D-FINE` or its methods in your work, please cite the following BibTeX entries:
|
|
|
|
| 16 |
|
| 17 |
This repository contains five trained PyTorch D-FINE models proposed in [D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement](https://arxiv.org/abs/2410.13842) by Yansong Peng, Hebei Li, Peixi Wu, Yueyi Zhang, Xiaoyan Sun, Feng Wu.
|
| 18 |
|
| 19 |
+
## D-FINE Variants
|
| 20 |
+
|
| 21 |
+
| Model | Params | mAP@[IoU=0.50:0.95] | Download |
|
| 22 |
+
|---------------|---------|--------------|----------|
|
| 23 |
+
| Nano | 3.76M | 0.825 | [d-fine-nano](https://huggingface.co/Laudando-Associates-LLC/d-fine-nano) |
|
| 24 |
+
| Small | 10.3M | 0.816 | [d-fine-small](https://huggingface.co/Laudando-Associates-LLC/d-fine-small) |
|
| 25 |
+
| Medium | 19.6M | 0.840 | [d-fine-medium](https://huggingface.co/Laudando-Associates-LLC/d-fine-medium) |
|
| 26 |
+
| Large | 31.2M | 0.828 | [d-fine-large](https://huggingface.co/Laudando-Associates-LLC/d-fine-large) |
|
| 27 |
+
| Extra Large | 62.7M | 0.803 | [d-fine-xlarge](https://huggingface.co/Laudando-Associates-LLC/d-fine-xlarge) |
|
| 28 |
+
|
| 29 |
|
| 30 |
## Citation
|
| 31 |
If you use `D-FINE` or its methods in your work, please cite the following BibTeX entries:
|