YOLO系列网络的改进仓库

# Multispectral-Object-Detection [](https://paperswithcode.com/sota/multispectral-object-detection-on-flir?p=cross-modality-fusion-transformer-for) [](https://paperswithcode.com/sota/pedestrian-detection-on-llvip?p=cross-modality-fusion-transformer-for) [](https://github.com/DocF/multispectral-object-detection/)  [](https://github.com/DocF/multispectral-object-detection) ## Intro Official Code for [Cross-Modality Fusion Transformer for Multispectral Object Detection](https://arxiv.org/abs/2111.00273). Multispectral Object Detection with Transformer and Yolov5 ## Abstract Multispectral image pairs can provide the combined information, making object detection applications more reliable and robust in the open world. To fully exploit the different modalities, we present a simple yet effective cross-modality feature fusion approach, named Cross-Modality Fusion Transformer (CFT) in this paper. Unlike prior CNNs-based works, guided by the Transformer scheme, our network learns long-range dependencies and integrates global contextual information in the feature extraction stage. More importantly, by leveraging the self attention of the Transformer, the network can naturally carry out simultaneous intra-modality and inter-modality fusion, and robustly capture the latent interactions between RGB and Thermal domains, thereby significantly improving the performance of multispectral object detection. Extensive experiments and ablation studies on multiple datasets demonstrate that our approach is effective and achieves state-of-the-art detection performance. ### Demo **Night Scene** <div align="left"> <img src="https://github.com/DocF/multispectral-object-detection/blob/main/video/demo1.gif" width="600"> </div> **Day Scene** <div align="left"> <img src="https://github.com/DocF/multispectral-object-detection/blob/main/video/demo.gif" width="600"> </div> ### Overview <div align="left"> <img src="https://github.com/DocF/multispectral-object-detection/blob/main/cft.png" width="800"> </div> ## Citation If you use this repo for your research, please cite our paper: ``` @article{fang2021cross, title={Cross-Modality Fusion Transformer for Multispectral Object Detection}, author={Fang Qingyun and Han Dapeng and Wang Zhaokui}, journal={arXiv preprint arXiv:2111.00273}, year={2021} } ``` ## Installation Python>=3.6.0 is required with all requirements.txt installed including PyTorch>=1.7 (The same as yolov5 https://github.com/ultralytics/yolov5 ). #### Clone the repo git clone https://github.com/DocF/multispectral-object-detection #### Install requirements ```bash $ cd multispectral-object-detection $ pip install -r requirements.txt ``` ## Dataset -[FLIR] [[Google Drive]](http://shorturl.at/ahAY4) [[Baidu Drive]](https://pan.baidu.com/s/1z2GHVD3WVlGsVzBR1ajSrQ?pwd=qwer) ```extraction code:qwer``` A new aligned version. -[LLVIP] [download](https://github.com/bupt-ai-cz/LLVIP) -[VEDAI] [download](https://downloads.greyc.fr/vedai/) You need to convert all annotations to YOLOv5 format. Refer: https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data ## Run #### Download the pretrained weights yolov5 weights (pre-train) -[yolov5s] [google drive](https://drive.google.com/file/d/1UGAsaOvV7jVrk0RvFVYL6Vq0K7NQLD8H/view?usp=sharing) -[yolov5m] [google drive](https://drive.google.com/file/d/1qB7L2vtlGppGjHp5xpXCKw14YHhbV4s1/view?usp=sharing) -[yolov5l] [google drive](https://drive.google.com/file/d/12OFGLF73CqTgOCMJAycZ8lB4eW19D0nb/view?usp=sharing) -[yolov5x] [google drive](https://drive.google.com/file/d/1e9xiQImx84KFQ_a7XXpn608I3rhRmKEn/view?usp=sharing) CFT weights -[LLVIP] [google drive](https://drive.google.com/file/d/18yLDUOxNXQ17oypQ-fAV9OS9DESOZQtV/view?usp=sharing) -[FLIR] [google drive](https://drive.google.com/file/d/1PwEOgT5ZOTjoKT2LpOzvCsxsVgwP8NIJ/view) #### Change the data cfg some example in data/multispectral/ #### Change the model cfg some example in models/transformer/ note!!! we used xxxx_transfomerx3_dataset.yaml in our paper. ### Train Test and Detect train: ``` python train.py``` test: ``` python test.py``` detect: ``` python detect_twostream.py``` ## Results |Dataset|CFT|mAP50|mAP75|mAP| |:---------: |------------|:-----:|:-----------------:|:-------------:| |FLIR||73.0|32.0|37.4| |FLIR| ✔️ |**78.7 (Δ5.7)**|**35.5 (Δ3.5)**|**40.2 (Δ2.8)**| |LLVIP||95.8|71.4|62.3| |LLVIP| ✔️ |**97.5 (Δ1.7)**|**72.9 (Δ1.5)**|**63.6 (Δ1.3)**| |VEDAI||79.7 | 47.7 | 46.8 |VEDAI| ✔️ |**85.3 (Δ5.6)**|**65.9(Δ18.2)**|**56.0 (Δ9.2)**| ### LLVIP Log Average Miss Rate |Model| Log Average Miss Rate | |:---------: |:--------------:| |YOLOv3-RGB|37.70%| |YOLOv3-IR|17.73%| |YOLOv5-RGB|22.59%| |YOLOv5-IR|10.66%| |Baseline(Ours)|**6.91%**| |CFT(Ours)|**5.40%**| Miss Rate - FPPI curve <div align="left"> <img src="https://github.com/DocF/multispectral-object-detection/blob/main/MR.png" width="500"> </div> #### References https://github.com/ultralytics/yolov5