https://github.com/ajinkya-kulkarni/d-fine

D-FINE: Redefine Regression Task of DETRs as Fine-grained Distribution Refinement [ICLR 2025 Spotlight]
Science Score: 10.0%

This score indicates how likely this project is to be science-related based on various indicators:
○
CITATION.cff file
○
codemeta.json file
○
.zenodo.json file
○
DOI references
✓
Academic publication links
Links to: arxiv.org
○
Academic email domains
○
Institutional organization owner
○
JOSS paper metadata
○
Scientific vocabulary similarity
Low similarity (6.6%) to scientific vocabulary
Last synced: 10 months ago · JSON representation
Repository

D-FINE: Redefine Regression Task of DETRs as Fine-grained Distribution Refinement [ICLR 2025 Spotlight]
Basic Info

Host: GitHub
Owner: ajinkya-kulkarni
License: apache-2.0
Language: Python
Default Branch: master
Homepage:
Size: 295 KB
Statistics

Stars: 0
Watchers: 0
Forks: 0
Open Issues: 0
Releases: 0
Fork of Peterande/D-FINE
Created over 1 year ago · Last pushed over 1 year ago
https://github.com/ajinkya-kulkarni/D-FINE/blob/master/



English | [](README_cn.md) | [](README_ja.md) | [English Blog](src/zoo/dfine/blog.md) | [](src/zoo/dfine/blog_cn.md)


  D-FINE: Redefine Regression Task of DETRs as Fine‑grained Distribution Refinement





    
        
    
    
        
    
    
        
    
    
        
    

      
        
    
    
        
    





     This is the official implementation of the paper:
    

    D-FINE: Redefine Regression Task of DETRs as Fine-grained Distribution Refinement





Yansong Peng, Hebei Li, Peixi Wu, Yueyi Zhang, Xiaoyan Sun, and Feng Wu



University of Science and Technology of China



    
        
    





If you like D-FINE, please give us a ! Your support motivates us to keep improving!



    


D-FINE is a powerful real-time object detector that redefines the bounding box regression task in DETRs as Fine-grained Distribution Refinement (FDR) and introduces Global Optimal Localization Self-Distillation (GO-LSD), achieving outstanding performance without introducing additional inference and training costs.


 Video 

We conduct object detection using D-FINE and YOLO11 on a complex street scene video from [YouTube](https://www.youtube.com/watch?v=CfhEWj9sd9A). Despite challenging conditions such as backlighting, motion blur, and dense crowds, D-FINE-X successfully detects nearly all targets, including subtle small objects like backpacks, bicycles, and traffic lights. Its confidence scores and the localization precision for blurred edges are significantly higher than those of YOLO11.



https://github.com/user-attachments/assets/e5933d8e-3c8a-400e-870b-4e452f5321d9



##  Updates
- [x] **\[2024.10.18\]** Release D-FINE series.
- [x] **\[2024.10.25\]** Add custom dataset finetuning configs ([#7](https://github.com/Peterande/D-FINE/issues/7)).
- [x] **\[2024.10.30\]** Update D-FINE-L (E25) pretrained model, with performance improved by 2.0%.
- [x] **\[2024.11.07\]** Release **D-FINE-N**, achiving 42.8% AP^val on COCO @ 472 FPS^T4!

## Model Zoo

### COCO
| Model | Dataset | AP^val | #Params | Latency | GFLOPs | config | checkpoint | logs |
| :---: | :---: | :---: |  :---: | :---: | :---: | :---: | :---: | :---: |
**D‑FINE‑N** | COCO | **42.8** | 4M | 2.12ms | 7 | [yml](./configs/dfine/dfine_hgnetv2_n_coco.yml) | [42.8](https://github.com/Peterande/storage/releases/download/dfinev1.0/dfine_n_coco.pth) | [url](https://raw.githubusercontent.com/Peterande/storage/refs/heads/master/logs/coco/dfine_n_coco_log.txt)
**D‑FINE‑S** | COCO | **48.5** | 10M | 3.49ms | 25 | [yml](./configs/dfine/dfine_hgnetv2_s_coco.yml) | [48.5](https://github.com/Peterande/storage/releases/download/dfinev1.0/dfine_s_coco.pth) | [url](https://raw.githubusercontent.com/Peterande/storage/refs/heads/master/logs/coco/dfine_s_coco_log.txt)
**D‑FINE‑M** | COCO | **52.3** | 19M | 5.62ms | 57 | [yml](./configs/dfine/dfine_hgnetv2_m_coco.yml) | [52.3](https://github.com/Peterande/storage/releases/download/dfinev1.0/dfine_m_coco.pth) | [url](https://raw.githubusercontent.com/Peterande/storage/refs/heads/master/logs/coco/dfine_m_coco_log.txt)
**D‑FINE‑L** | COCO | **54.0** | 31M | 8.07ms | 91 | [yml](./configs/dfine/dfine_hgnetv2_l_coco.yml) | [54.0](https://github.com/Peterande/storage/releases/download/dfinev1.0/dfine_l_coco.pth) | [url](https://raw.githubusercontent.com/Peterande/storage/refs/heads/master/logs/coco/dfine_l_coco_log.txt)
**D‑FINE‑X** | COCO | **55.8** | 62M | 12.89ms | 202 | [yml](./configs/dfine/dfine_hgnetv2_x_coco.yml) | [55.8](https://github.com/Peterande/storage/releases/download/dfinev1.0/dfine_x_coco.pth) | [url](https://raw.githubusercontent.com/Peterande/storage/refs/heads/master/logs/coco/dfine_x_coco_log.txt)


### Objects365+COCO
| Model | Dataset | AP^val | #Params | Latency | GFLOPs | config | checkpoint | logs |
| :---: | :---: | :---: |  :---: | :---: | :---: | :---: | :---: | :---: |
**D‑FINE‑S** | Objects365+COCO | **50.7** | 10M | 3.49ms | 25 | [yml](./configs/dfine/objects365/dfine_hgnetv2_s_obj2coco.yml) | [50.7](https://github.com/Peterande/storage/releases/download/dfinev1.0/dfine_s_obj2coco.pth) | [url](https://raw.githubusercontent.com/Peterande/storage/refs/heads/master/logs/obj2coco/dfine_s_obj2coco_log.txt)
**D‑FINE‑M** | Objects365+COCO | **55.1** | 19M | 5.62ms | 57 | [yml](./configs/dfine/objects365/dfine_hgnetv2_m_obj2coco.yml) | [55.1](https://github.com/Peterande/storage/releases/download/dfinev1.0/dfine_m_obj2coco.pth) | [url](https://raw.githubusercontent.com/Peterande/storage/refs/heads/master/logs/obj2coco/dfine_m_obj2coco_log.txt)
**D‑FINE‑L** | Objects365+COCO | **57.3** | 31M | 8.07ms | 91 | [yml](./configs/dfine/objects365/dfine_hgnetv2_l_obj2coco.yml) | [57.3](https://github.com/Peterande/storage/releases/download/dfinev1.0/dfine_l_obj2coco_e25.pth) | [url](https://raw.githubusercontent.com/Peterande/storage/refs/heads/master/logs/obj2coco/dfine_l_obj2coco_log_e25.txt)
**D‑FINE‑X** | Objects365+COCO | **59.3** | 62M | 12.89ms | 202 | [yml](./configs/dfine/objects365/dfine_hgnetv2_x_obj2coco.yml) | [59.3](https://github.com/Peterande/storage/releases/download/dfinev1.0/dfine_x_obj2coco.pth) | [url](https://raw.githubusercontent.com/Peterande/storage/refs/heads/master/logs/obj2coco/dfine_x_obj2coco_log.txt)

**We highly recommend that you use the Objects365 pre-trained model for fine-tuning:**

 **Important**: Please note that this is generally beneficial for complex scene understanding. If your categories are very simple, it might lead to overfitting and suboptimal performance.

  Pretrained Models on Objects365 (Best generalization) 

| Model | Dataset | AP^val | AP⁵⁰⁰⁰ | #Params | Latency | GFLOPs | config | checkpoint | logs |
| :---: | :---: | :---: |  :---: | :---: | :---: | :---: | :---: | :---: | :---: |
**D‑FINE‑S** | Objects365 | **31.0** | **30.5** | 10M | 3.49ms | 25 | [yml](./configs/dfine/objects365/dfine_hgnetv2_s_obj365.yml) | [30.5](https://github.com/Peterande/storage/releases/download/dfinev1.0/dfine_s_obj365.pth) | [url](https://raw.githubusercontent.com/Peterande/storage/refs/heads/master/logs/obj365/dfine_s_obj365_log.txt)
**D‑FINE‑M** | Objects365 | **38.6** | **37.4** | 19M | 5.62ms | 57 | [yml](./configs/dfine/objects365/dfine_hgnetv2_m_obj365.yml) | [37.4](https://github.com/Peterande/storage/releases/download/dfinev1.0/dfine_m_obj365.pth) | [url](https://raw.githubusercontent.com/Peterande/storage/refs/heads/master/logs/obj365/dfine_m_obj365_log.txt)
**D‑FINE‑L** | Objects365 | - | **40.6** | 31M | 8.07ms | 91 | [yml](./configs/dfine/objects365/dfine_hgnetv2_l_obj365.yml) | [40.6](https://github.com/Peterande/storage/releases/download/dfinev1.0/dfine_l_obj365.pth) | [url](https://raw.githubusercontent.com/Peterande/storage/refs/heads/master/logs/obj365/dfine_l_obj365_log.txt)
**D‑FINE‑L (E25)** | Objects365 | **44.7** | **42.6** | 31M | 8.07ms | 91 | [yml](./configs/dfine/objects365/dfine_hgnetv2_l_obj365.yml) | [42.6](https://github.com/Peterande/storage/releases/download/dfinev1.0/dfine_l_obj365_e25.pth) | [url](https://raw.githubusercontent.com/Peterande/storage/refs/heads/master/logs/obj365/dfine_l_obj365_log_e25.txt)
**D‑FINE‑X** | Objects365 | **49.5** | **46.5** | 62M | 12.89ms | 202 | [yml](./configs/dfine/objects365/dfine_hgnetv2_x_obj365.yml) | [46.5](https://github.com/Peterande/storage/releases/download/dfinev1.0/dfine_x_obj365.pth) | [url](https://raw.githubusercontent.com/Peterande/storage/refs/heads/master/logs/obj365/dfine_x_obj365_log.txt)
- **E25**: Re-trained and extended the pretraining to 25 epochs.
- **AP^val** is evaluated on *Objects365* full validation set.
- **AP⁵⁰⁰⁰** is evaluated on the first 5000 samples of the *Objects365* validation set.


**Notes:**
- **AP^val** is evaluated on *MSCOCO val2017* dataset.
- **Latency** is evaluated on a single T4 GPU with $batch\\_size = 1$, $fp16$, and $TensorRT==10.4.0$.
- **Objects365+COCO** means finetuned model on *COCO* using pretrained weights trained on *Objects365*.



## Quick start

### Setup

```shell
conda create -n dfine python=3.11.9
conda activate dfine
pip install -r requirements.txt
```


### Data Preparation


 COCO2017 Dataset 

1. Download COCO2017 from [OpenDataLab](https://opendatalab.com/OpenDataLab/COCO_2017) or [COCO](https://cocodataset.org/#download).
1. Modify paths in [coco_detection.yml](./configs/dataset/coco_detection.yml)

    ```yaml
    train_dataloader:
        img_folder: /data/COCO2017/train2017/
        ann_file: /data/COCO2017/annotations/instances_train2017.json
    val_dataloader:
        img_folder: /data/COCO2017/val2017/
        ann_file: /data/COCO2017/annotations/instances_val2017.json
    ```




 Objects365 Dataset 

1. Download Objects365 from [OpenDataLab](https://opendatalab.com/OpenDataLab/Objects365).

2. Set the Base Directory:
```shell
export BASE_DIR=/data/Objects365/data
```

3. Extract and organize the downloaded files, resulting directory structure:

```shell
${BASE_DIR}/train
 images
    v1
       patch0
          000000000.jpg
          000000001.jpg
          ... (more images)
    v2
       patchx
          000000000.jpg
          000000001.jpg
          ... (more images)
 zhiyuan_objv2_train.json
```

```shell
${BASE_DIR}/val
 images
    v1
       patch0
          000000000.jpg
          ... (more images)
    v2
       patchx
          000000000.jpg
          ... (more images)
 zhiyuan_objv2_val.json
```

4. Create a New Directory to Store Images from the Validation Set:
```shell
mkdir -p ${BASE_DIR}/train/images_from_val
```

5. Copy the v1 and v2 folders from the val directory into the train/images_from_val directory
```shell
cp -r ${BASE_DIR}/val/images/v1 ${BASE_DIR}/train/images_from_val/
cp -r ${BASE_DIR}/val/images/v2 ${BASE_DIR}/train/images_from_val/
```

6. Run remap_obj365.py to merge a subset of the validation set into the training set. Specifically, this script moves samples with indices between 5000 and 800000 from the validation set to the training set.
```shell
python tools/remap_obj365.py --base_dir ${BASE_DIR}
```


7. Run the resize_obj365.py script to resize any images in the dataset where the maximum edge length exceeds 640 pixels. Use the updated JSON file generated in Step 5 to process the sample data. Ensure that you resize images in both the train and val datasets to maintain consistency.
```shell
python tools/resize_obj365.py --base_dir ${BASE_DIR}
```

8. Modify paths in [obj365_detection.yml](./configs/dataset/obj365_detection.yml)

    ```yaml
    train_dataloader:
        img_folder: /data/Objects365/data/train
        ann_file: /data/Objects365/data/train/new_zhiyuan_objv2_train_resized.json
    val_dataloader:
        img_folder: /data/Objects365/data/val/
        ann_file: /data/Objects365/data/val/new_zhiyuan_objv2_val_resized.json
    ```





CrowdHuman

Download COCO format dataset here: [url](https://aistudio.baidu.com/datasetdetail/231455)




Custom Dataset

To train on your custom dataset, you need to organize it in the COCO format. Follow the steps below to prepare your dataset:

1. **Set `remap_mscoco_category` to `False`:**

    This prevents the automatic remapping of category IDs to match the MSCOCO categories.

    ```yaml
    remap_mscoco_category: False
    ```

2. **Organize Images:**

    Structure your dataset directories as follows:

    ```shell
    dataset/
     images/
        train/
           image1.jpg
           image2.jpg
           ...
        val/
           image1.jpg
           image2.jpg
           ...
     annotations/
         instances_train.json
         instances_val.json
         ...
    ```

    - **`images/train/`**: Contains all training images.
    - **`images/val/`**: Contains all validation images.
    - **`annotations/`**: Contains COCO-formatted annotation files.

3. **Convert Annotations to COCO Format:**

    If your annotations are not already in COCO format, you'll need to convert them. You can use the following Python script as a reference or utilize existing tools:

    ```python
    import json

    def convert_to_coco(input_annotations, output_annotations):
        # Implement conversion logic here
        pass

    if __name__ == "__main__":
        convert_to_coco('path/to/your_annotations.json', 'dataset/annotations/instances_train.json')
    ```

4. **Update Configuration Files:**

    Modify your [custom_detection.yml](./configs/dataset/custom_detection.yml).

    ```yaml
    task: detection

    evaluator:
      type: CocoEvaluator
      iou_types: ['bbox', ]

    num_classes: 777 # your dataset classes
    remap_mscoco_category: False

    train_dataloader:
      type: DataLoader
      dataset:
        type: CocoDetection
        img_folder: /data/yourdataset/train
        ann_file: /data/yourdataset/train/train.json
        return_masks: False
        transforms:
          type: Compose
          ops: ~
      shuffle: True
      num_workers: 4
      drop_last: True
      collate_fn:
        type: BatchImageCollateFunction

    val_dataloader:
      type: DataLoader
      dataset:
        type: CocoDetection
        img_folder: /data/yourdataset/val
        ann_file: /data/yourdataset/val/ann.json
        return_masks: False
        transforms:
          type: Compose
          ops: ~
      shuffle: False
      num_workers: 4
      drop_last: False
      collate_fn:
        type: BatchImageCollateFunction
    ```




## Usage

 COCO2017 


1. Set Model
```shell
export model=l  # n s m l x
```

2. Training
```shell
CUDA_VISIBLE_DEVICES=0,1,2,3 torchrun --master_port=7777 --nproc_per_node=4 train.py -c configs/dfine/dfine_hgnetv2_${model}_coco.yml --use-amp --seed=0
```


3. Testing
```shell
CUDA_VISIBLE_DEVICES=0,1,2,3 torchrun --master_port=7777 --nproc_per_node=4 train.py -c configs/dfine/dfine_hgnetv2_${model}_coco.yml --test-only -r model.pth
```


4. Tuning
```shell
CUDA_VISIBLE_DEVICES=0,1,2,3 torchrun --master_port=7777 --nproc_per_node=4 train.py -c configs/dfine/dfine_hgnetv2_${model}_coco.yml --use-amp --seed=0 -t model.pth
```




 Objects365 to COCO2017 

1. Set Model
```shell
export model=l  # n s m l x
```

2. Training on Objects365
```shell
CUDA_VISIBLE_DEVICES=0,1,2,3 torchrun --master_port=7777 --nproc_per_node=4 train.py -c configs/dfine/objects365/dfine_hgnetv2_${model}_obj365.yml --use-amp --seed=0
```

3. Tuning on COCO2017
```shell
CUDA_VISIBLE_DEVICES=0,1,2,3 torchrun --master_port=7777 --nproc_per_node=4 train.py -c configs/dfine/objects365/dfine_hgnetv2_${model}_obj2coco.yml --use-amp --seed=0 -t model.pth
```


4. Testing
```shell
CUDA_VISIBLE_DEVICES=0,1,2,3 torchrun --master_port=7777 --nproc_per_node=4 train.py -c configs/dfine/dfine_hgnetv2_${model}_coco.yml --test-only -r model.pth
```




 Custom Dataset 

1. Set Model
```shell
export model=l  # n s m l x
```

2. Training on Custom Dataset
```shell
CUDA_VISIBLE_DEVICES=0,1,2,3 torchrun --master_port=7777 --nproc_per_node=4 train.py -c configs/dfine/custom/dfine_hgnetv2_${model}_custom.yml --use-amp --seed=0
```

3. Testing
```shell
CUDA_VISIBLE_DEVICES=0,1,2,3 torchrun --master_port=7777 --nproc_per_node=4 train.py -c configs/dfine/custom/dfine_hgnetv2_${model}_custom.yml --test-only -r model.pth
```

4. Tuning on Custom Dataset
```shell
CUDA_VISIBLE_DEVICES=0,1,2,3 torchrun --master_port=7777 --nproc_per_node=4 train.py -c configs/dfine/custom/objects365/dfine_hgnetv2_${model}_obj2custom.yml --use-amp --seed=0 -t model.pth
```

5. **[Optional]** Modify Class Mappings:

When using the Objects365 pre-trained weights to train on your custom dataset, the example assumes that your dataset only contains the classes `'Person'` and `'Car'`. For faster convergence, you can modify `self.obj365_ids` in `src/solver/_solver.py` as follows:


```python
self.obj365_ids = [0, 5]  # Person, Cars
```
You can replace these with any corresponding classes from your dataset. The list of Objects365 classes with their corresponding IDs:
https://github.com/Peterande/D-FINE/blob/352a94ece291e26e1957df81277bef00fe88a8e3/src/solver/_solver.py#L330

New training command:

```shell
CUDA_VISIBLE_DEVICES=0,1,2,3 torchrun --master_port=7777 --nproc_per_node=4 train.py -c configs/dfine/custom/dfine_hgnetv2_${model}_custom.yml --use-amp --seed=0 -t model.pth
```

However, if you don't wish to modify the class mappings, the pre-trained Objects365 weights will still work without any changes. Modifying the class mappings is optional and can potentially accelerate convergence for specific tasks.






 Customizing Batch Size 

For example, if you want to double the total batch size when training D-FINE-L on COCO2017, here are the steps you should follow:

1. **Modify your [dataloader.yml](./configs/dfine/include/dataloader.yml)** to increase the `total_batch_size`:

    ```yaml
    train_dataloader:
        total_batch_size: 64  # Previously it was 32, now doubled
    ```

2. **Modify your [dfine_hgnetv2_l_coco.yml](./configs/dfine/dfine_hgnetv2_l_coco.yml)**. Heres how the key parameters should be adjusted:

    ```yaml
    optimizer:
    type: AdamW
    params:
        -
        params: '^(?=.*backbone)(?!.*norm|bn).*$'
        lr: 0.000025  # doubled, linear scaling law
        -
        params: '^(?=.*(?:encoder|decoder))(?=.*(?:norm|bn)).*$'
        weight_decay: 0.

    lr: 0.0005  # doubled, linear scaling law
    betas: [0.9, 0.999]
    weight_decay: 0.0001  # need a grid search

    ema:  # added EMA settings
        decay: 0.9998  # adjusted by 1 - (1 - decay) * 2
        warmups: 500  # halved

    lr_warmup_scheduler:
        warmup_duration: 250  # halved
    ```





 Customizing Input Size 

If you'd like to train **D-FINE-L** on COCO2017 with an input size of 320x320, follow these steps:

1. **Modify your [dataloader.yml](./configs/dfine/include/dataloader.yml)**:

    ```yaml

    train_dataloader:
    dataset:
        transforms:
            ops:
                - {type: Resize, size: [320, 320], }
    collate_fn:
        base_size: 320
    dataset:
        transforms:
            ops:
                - {type: Resize, size: [320, 320], }
    ```

2. **Modify your [dfine_hgnetv2.yml](./configs/dfine/include/dfine_hgnetv2.yml)**:

    ```yaml
    eval_spatial_size: [320, 320]
    ```



## Tools

 Deployment 


1. Setup
```shell
pip install onnx onnxsim
export model=l  # n s m l x
```

2. Export onnx
```shell
python tools/deployment/export_onnx.py --check -c configs/dfine/dfine_hgnetv2_${model}_coco.yml -r model.pth
```

3. Export [tensorrt](https://docs.nvidia.com/deeplearning/tensorrt/install-guide/index.html)
```shell
trtexec --onnx="model.onnx" --saveEngine="model.engine" --fp16
```




 Inference (Visualization) 


1. Setup
```shell
pip install -r tools/inference/requirements.txt
export model=l  # n s m l x
```



2. Inference (onnxruntime / tensorrt / torch)

Inference on images and videos is now supported.
```shell
python tools/inference/onnx_inf.py --onnx model.onnx --input image.jpg  # video.mp4
python tools/inference/trt_inf.py --trt model.engine --input image.jpg
python tools/inference/torch_inf.py -c configs/dfine/dfine_hgnetv2_${model}_coco.yml -r model.pth --input image.jpg --device cuda:0
```



 Benchmark 

1. Setup
```shell
pip install -r tools/benchmark/requirements.txt
export model=l  # n s m l x
```


2. Model FLOPs, MACs, and Params
```shell
python tools/benchmark/get_info.py -c configs/dfine/dfine_hgnetv2_${model}_coco.yml
```

2. TensorRT Latency
```shell
python tools/benchmark/trt_benchmark.py --COCO_dir path/to/COCO2017 --engine_dir model.engine
```



 Fiftyone Visualization  

1. Setup
```shell
pip install fiftyone
export model=l  # n s m l x
```
4. Voxel51 Fiftyone Visualization ([fiftyone](https://github.com/voxel51/fiftyone))
```shell
python tools/visualization/fiftyone_vis.py -c configs/dfine/dfine_hgnetv2_${model}_coco.yml -r model.pth
```



 Others 

1. Auto Resume Training
```shell
bash reference/safe_training.sh
```

2. Converting Model Weights
```shell
python reference/convert_weight.py model.pth
```


## Figures and Visualizations


 FDR and GO-LSD 

1. Overview of D-FINE with FDR. The probability distributions that act as a more fine-
grained intermediate representation are iteratively refined by the decoder layers in a residual manner.
Non-uniform weighting functions are applied to allow for finer localization.


    


2. Overview of GO-LSD process. Localization knowledge from the final layers refined
distributions is distilled into earlier layers through DDF loss with decoupled weighting strategies.


    





 Distributions 

Visualizations of FDR across detection scenarios with initial and refined bounding boxes, along with unweighted and weighted distributions.


    





 Hard Cases 

The following visualization demonstrates D-FINE's predictions in various complex detection scenarios. These include cases with occlusion, low-light conditions, motion blur, depth of field effects, and densely populated scenes. Despite these challenges, D-FINE consistently produces accurate localization results.


    










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

 bibtex 

```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}
}
```


## Acknowledgement
Our work is built upon [RT-DETR](https://github.com/lyuwenyu/RT-DETR).
Thanks to the inspirations from [RT-DETR](https://github.com/lyuwenyu/RT-DETR), [GFocal](https://github.com/implus/GFocal), [LD](https://github.com/HikariTJU/LD), and [YOLOv9](https://github.com/WongKinYiu/yolov9).

 Feel free to contribute and reach out if you have any questions!