This is the official repository with PyTorch implementation of LW-DETR: A Transformer Replacement to YOLO for Real-Time Detection.
☀️ If you find this work useful for your research, please kindly star our repo and cite our paper! ☀️
Release a series of real-time detection models in LW-DETR, including LW-DETR-tiny, LW-DETR-small, LW-DETR-medium, LW-DETR-large and LW-DETR-xlarge, named
Release a series of pretrained models in LW-DETR. Please refer to Hugging Face to download.
[2024/7/15] We present OVLW-DETR, an efficient open-vocabulary detector with outstanding performance and low latency, built upon LW-DETR. It surpasses existing real-time open-vocabulary detectors on the standard Zero-Shot LVIS benchmark. The source code and pre-trained model is comming soon, please stay tuned!
1. Introduction
2. Installation
3. Preparation
4. Train
5. Eval
6. Deploy
7. Main Results
8. References
9. Citation
LW-DETR is a light-weight detection tranformer, which outperforms YOLOs for real-time object detection. The architecture is a simple stack of a ViT encoder, a projector, and a shallow DETR decoder. LW-DETR leverages recent advanced techniques, such as training-effective techniques, e.g., improved loss and pretraining, and interleaved window and global attentions for reducing the ViT encoder complexity. LW-DETR improves the ViT encoder by aggregating multi-level feature maps, and the intermediate and final feature maps in the ViT encoder, forming richer feature maps, and introduces window-major feature map organization for improving the efficiency of interleaved attention computation. LW-DETR achieves superior performance than on existing real-time detectors, e.g., YOLO and its variants, on COCO and other benchmark datasets.
The code is developed and validated under python=3.8.19, pytorch=1.13.0, cuda=11.6,TensorRT-8.6.1.6. Higher versions might be available as well.
Create your own Python environment with Anaconda.
conda create -n lwdetr python=3.8.19 conda activate lwdetr
Clone this repo.
git clone https://github.com/Atten4Vis/LW-DETR.gitcd LW-DETR
Install PyTorch and torchvision.
Follow the instruction on https://pytorch.org/get-started/locally/.
# an example:conda install pytorch==1.13.0 torchvision==0.14.0 pytorch-cuda=11.6 -c pytorch -c nvidia
Install required packages.
For training and evaluation:
pip install -r requirements.txt
For deployment:
Please refer to NVIDIA for installation instruction of TensorRT
pip install -r deploy/requirements.txt
Compiling CUDA operators
cd models/ops python setup.py build install# unit test (should see all checking is True)python test.pycd ../..
For MS COCO dataset, please download and extract COCO 2017 train and val images with annotations from http://cocodataset.org. We expect the directory structure to be the following:
COCODIR/ ├── train2017/ ├── val2017/ └── annotations/ ├── instances_train2017.json └── instances_val2017.json
For Objects365 dataset for pretraining, please download Objects365 images with annotations from https://www.objects365.org/overview.html.
All the checkpoints can be found in Hugging Face.
Pretraining on Objects365.
Pretrained the ViT.
We pretrain the ViT on the dataset Objects365 using a MIM method, CAE v2, based on the pretrained models. Please refer to the following link to download the pretrained models, and put them into pretrain_weights/.
| Model | Comment |
|---|---|
| caev2_tiny_300e_objects365 | pretrained ViT model on objects365 for LW-DETR-tiny/small using CAE v2 |
| caev2_tiny_300e_objects365 | pretrained ViT model on objects365 for LW-DETR-medium/large using CAE v2 |
| caev2_tiny_300e_objects365 | pretrained ViT model on objects365 for LW-DETR-xlarge using CAE v2 |
Pretrained LW-DETR.
We retrain the encoder and train the projector and the decoder on Objects365 in a supervision manner. Please refer to the following link to download the pretrained models, and put them into pretrain_weights/.
| Model | Comment |
|---|---|
| LWDETR_tiny_30e_objects365 | pretrained LW-DETR-tiny model on objects365 |
| LWDETR_small_30e_objects365 | pretrained LW-DETR-small model on objects365 |
| LWDETR_medium_30e_objects365 | pretrained LW-DETR-medium model on objects365 |
| LWDETR_large_30e_objects365 | pretrained LW-DETR-large model on objects365 |
| LWDETR_xlarge_30e_objects365 | pretrained LW-DETR-xlarge model on objects365 |
Finetuning on COCO.
We finetune the pretrained model on COCO. If you want to reimplement our repo, please skip this step. If you want to directly evaluate our trained models, please refer to the following link to download the finetuned models, and put them into output/.
| Model | Comment |
|---|---|
| LWDETR_tiny_60e_coco | finetuned LW-DETR-tiny model on COCO |
| LWDETR_small_60e_coco | finetuned LW-DETR-small model on COCO |
| LWDETR_medium_60e_coco | finetuned LW-DETR-medium model on COCO |
| LWDETR_large_60e_coco | finetuned LW-DETR-large model on COCO |
| LWDETR_xlarge_60e_coco | finetuned LW-DETR-xlarge model on COCO |
You can directly run scripts/lwdetr_ file for the training process on coco dataset.
sh scripts/lwdetr_tiny_coco_train.sh /path/to/your/COCODIR
sh scripts/lwdetr_small_coco_train.sh /path/to/your/COCODIR
sh scripts/lwdetr_medium_coco_train.sh /path/to/your/COCODIR
sh scripts/lwdetr_large_coco_train.sh /path/to/your/COCODIR
sh scripts/lwdetr_xlarge_coco_train.sh /path/to/your/COCODIR
You can directly run scripts/lwdetr_ file for the evaluation process on coco dataset. Please refer to 3. Preparation to download a series of LW-DETR models.
sh scripts/lwdetr_tiny_coco_eval.sh /path/to/your/COCODIR /path/to/your/checkpoint
sh scripts/lwdetr_small_coco_eval.sh /path/to/your/COCODIR /path/to/your/checkpoint
sh scripts/lwdetr_medium_coco_eval.sh /path/to/your/COCODIR /path/to/your/checkpoint
sh scripts/lwdetr_large_coco_eval.sh /path/to/your/COCODIR /path/to/your/checkpoint
sh scripts/lwdetr_xlarge_coco_eval.sh /path/to/your/COCODIR /path/to/your/checkpoint
You can run scripts/lwdetr_ file to export models for development. Before execution, please ensure that TensorRT and cuDNN environment variables are correctly set.
# export ONNX modelsh scripts/lwdetr_tiny_coco_export.sh /path/to/your/COCODIR /path/to/your/checkpoint# convert model from ONNX to TensorRT engine as wellsh scripts/lwdetr_tiny_coco_export.sh /path/to/your/COCODIR /path/to/your/checkpoint --trt
# export ONNX modelsh scripts/lwdetr_small_coco_export.sh /path/to/your/COCODIR /path/to/your/checkpoint# convert model from ONNX to TensorRT engine as wellsh scripts/lwdetr_small_coco_export.sh /path/to/your/COCODIR /path/to/your/checkpoint --trt
# export ONNX modelsh scripts/lwdetr_medium_coco_export.sh /path/to/your/COCODIR /path/to/your/checkpoint# convert model from ONNX to TensorRT engine as wellsh scripts/lwdetr_medium_coco_export.sh /path/to/your/COCODIR /path/to/your/checkpoint --trt
# export ONNX modelsh scripts/lwdetr_large_coco_export.sh /path/to/your/COCODIR /path/to/your/checkpoint# convert model from ONNX to TensorRT engine as wellsh scripts/lwdetr_large_coco_export.sh /path/to/your/COCODIR /path/to/your/checkpoint --trt
# export ONNX modelsh scripts/lwdetr_xlarge_coco_export.sh /path/to/your/COCODIR /path/to/your/checkpoint# convert model from ONNX to TensorRT engine as wellsh scripts/lwdetr_xlarge_coco_export.sh /path/to/your/COCODIR /path/to/your/checkpoint --trt
You can use deploy/benchmark.py tool to run benchmarks of inference latency.
# evaluate and benchmark the latency on a onnx modelpython deploy/benchmark.py --path=/path/to/your/onnxmodel --coco_path=/path/to/your/COCODIR --run_benchmark # evaluate and benchmark the latency on a TensorRT enginepython deploy/benchmark.py --path=/path/to/your/trtengine --coco_path=/path/to/your/COCODIR --run_benchmark
The main results on coco dataset. We report the mAP as reported in the original paper, as well as the mAP obtained from re-implementation.
Method | pretraining | Params (M) | FLOPs (G) | Model Latency (ms) | Total Latency (ms) | mAP | Download |
|---|---|---|---|---|---|---|---|
LW-DETR-tiny | ✔ | 12.1 | 11.2 | 2.0 | 2.0 | 42.6(42.9) | Link |
LW-DETR-small | ✔ | 14.6 | 16.6 | 2.9 | 2.9 | 48.0(48.1) | Link |
LW-DETR-medium | ✔ | 28.2 | 42.8 | 5.6 | 5.6 | 52.5(52.6) | Link |
LW-DETR-large | ✔ | 46.8 | 71.6 | 8.8 | 8.8 | 56.1(56.1) | Link |
LW-DETR-xlarge | ✔ | 118.0 | 174.2 | 19.1 | 19.1 | 58.3(58.3) | Link |
Our project is conducted based on the following public paper with code:
Group DETR
ViTDet
Deformable DETR
Conditional DETR
YOLOv8
AlignDETR
CAE v2
CAE
If you find this code useful in your research, please kindly consider citing our paper:
@article{chen2024lw,title={LW-DETR: A Transformer Replacement to YOLO for Real-Time Detection},author={Chen, Qiang and Su, Xiangbo and Zhang, Xinyu and Wang, Jian and Chen, Jiahui and Shen, Yunpeng and Han, Chuchu and Chen, Ziliang and Xu, Weixiang and Li, Fanrong and others},journal={arXiv preprint arXiv:2406.03459},year={2024}} 
