README.md

---
language:
  - en
license: apache-2.0
tags:
    - object-detection
    - AgTech
    - transformers
library_name: pytorch
inference: false
datasets:
  - Laudando-Associates-LLC/pucks
base_model: Laudando-Associates-LLC/d-fine
base_model_relation: finetune
model-index:
- name: D-FINE Medium
  results:
  - task:
      type: object-detection
      name: Object Detection
    dataset:
      type: Laudando-Associates-LLC/pucks
      name: L&A Pucks Dataset
      config: default
      split: validation
    metrics:
    - type: mean_average_precision
      name: mAP@[IoU=0.50:0.95]
      value: 0.840
    - type: mean_average_precision
      name: mAP@0.50
      value: 0.992
    - type: mean_average_precision
      name: mAP@0.75
      value: 0.974
    - type: recall
      name: AR@[IoU=0.50:0.95 | maxDets=100]
      value: 0.894
    - type: recall
      name: AR@0.50
      value: 1.000
    - type: recall
      name: AR@0.75
      value: 0.988
    - type: f1
      value: 0.924
    - type: precision
      value: 0.898
    - type: recall
      value: 0.952
    - type: iou
      value: 0.784
---

<h1 align="center"><strong>D-FINE Medium</strong></h1>

<p align="center">
  <a href="https://huggingface.co/Laudando-Associates-LLC/d-fine-medium">
    <img src="https://img.shields.io/badge/HuggingFace-Model-yellow?logo=huggingface&style=for-the-badge">
  </a>
</p>

This repository contains the [D-FINE](https://arxiv.org/abs/2410.13842) Medium model, a real-time object detector designed for efficient and accurate object detection tasks.

<p align="center">
  <img src="assets/medium.png" alt="Medium Detections" />
</p>

## Try it in the Browser

You can test this model using our interactive Gradio demo:

<p align="center">
  <a href="https://huggingface.co/spaces/Laudando-Associates-LLC/d-fine-demo">
    <img src="https://img.shields.io/badge/Launch%20Demo-Gradio-FF4B4B?logo=gradio&logoColor=white&style=for-the-badge">
  </a>
</p>

## Model Overview

* Architecture: D-FINE Medium

* Parameters: 19.6M

* Performance: 
  
  - mAP@[0.50:0.95]: 0.840

  - mAP@[0.50]: 0.992

  - AR@[0.50:0.95]: 0.894

  - F1 Score: 0.924

* Framework: PyTorch / ONNX

* Training Hardware: 2× NVIDIA RTX A6000 GPUs

## Download

| Format | Link |
|:--------:|:------:|
| ONNX   | <a href="https://huggingface.co/Laudando-Associates-LLC/d-fine-medium/resolve/main/model.onnx"><img src="https://img.shields.io/badge/-ONNX-005CED?style=for-the-badge&logo=onnx&logoColor=white"></a> |
| PyTorch | <a href="https://huggingface.co/Laudando-Associates-LLC/d-fine-medium/resolve/main/pytorch_model.bin"><img src="https://img.shields.io/badge/PyTorch-EE4C2C?style=for-the-badge&logo=pytorch&logoColor=white"></a> |

## Usage

To utilize this model, ensure you have the shared [D-FINE processor](https://huggingface.co/Laudando-Associates-LLC/d-fine):

```python
from transformers import AutoProcessor, AutoModel

# Load processor
processor = AutoProcessor.from_pretrained("Laudando-Associates-LLC/d-fine", trust_remote_code=True)

# Load model
model = AutoModel.from_pretrained("Laudando-Associates-LLC/d-fine-medium", trust_remote_code=True)

# Process image
inputs = processor(image)

# Run inference
outputs = model(**inputs, conf_threshold=0.4)
```

## Evaluation

This model was trained and evaluated on the [L&A Pucks Dataset](https://huggingface.co/datasets/Laudando-Associates-LLC/pucks).

## License

This model is licensed under the [Apache License 2.0](https://github.com/Peterande/D-FINE/blob/master/LICENSE).

## Citation
If you use `D-FINE` or its methods in your work, please cite the following BibTeX entries:

```latex
@misc{peng2024dfine,
      title={D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement},
      author={Yansong Peng and Hebei Li and Peixi Wu and Yueyi Zhang and Xiaoyan Sun and Feng Wu},
      year={2024},
      eprint={2410.13842},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}
```