Based on YOLOAir🔥🔥🔥 
The YOLOAir2 algorithm library is a PyTorch-based combination toolbox for the YOLO series of algorithms. Unified model code framework, unified application, unified improvement, easy module combination, and building a more powerful network model.
English | 简体中文
Main features🚀 • Use 🍉 • document📒 • report a problem🌟 • discuss✌️ • Effect preview🚀
☁️💡🎈YOLOAir2 is the second version of the YOLOAir series, The framework is based on YOLOv7, including YOLOv7, YOLOv6, YOLOv5, YOLOX, YOLOR, YOLOv4, YOLOv3, Transformer, Attention and Improved-YOLOv7... Support to improve Backbone, Neck, Head, Loss, IoU, NMS and other modules, As a perfection and addition of YOLOAir
Model diversification: Build different detection network models based on different network modules.
Modular componentization: Help users to customize and quickly combine Backbone, Neck, and Head to diversify network models, help scientific research improve detection algorithms, model improvement, and network arrangement and combination🏆. Build powerful network models.
Unified model code framework, unified application method, unified parameter adjustment, unified improvement, integrated multi-task, easy module combination, and building a more powerful network model.
Built-in integration YOLOv5, YOLOv7, YOLOv6, YOLOX, YOLOR, Transformer, PP-YOLO, PP-YOLOv2, PP-YOLOE, PP-YOLOEPlus, Scaled_YOLOv4, YOLOv3, YOLOv4, YOLO-Face, TPH-YOLO, YOLOv5Lite, SPD-YOLO, SlimNeck-YOLO, PicoDet and other model network structures... Integrate multiple detection algorithms and related multi-task models Use a unified model code framework, integrated in the YOLOAir library, unified application method. It is convenient for researchers to improve the algorithm model of the paper, compare the models, and realize the diversification of network combinations. Contains lightweight models and models with higher precision, reasonably selected according to the scene, and strikes a balance between precision and speed. At the same time, the library supports the decoupling of different structures and module components, allowing the modules to be componentized. By combining different module components, users can customize and build different detection models according to different data sets or different business scenarios.
Supports integrated multi-tasks, including target detection, instance segmentation, image classification, pose estimation, face detection, target tracking and other tasks
project address🌟: https://github.com/iscyy/yoloair
🚀Support more YOLO series algorithm model improvements (continuously updated...)
The YOLOAir algorithm library summarizes a variety of mainstream YOLO series detection models, and a set of codes integrates multiple models:
- Built-in integrated YOLOv5 model network structure, YOLOv7 model network structure, YOLOv6 model network structure, PP-YOLO model network structure, PP-YOLOE model network structure, PP-YOLOEPlus model network structure, YOLOR model network structure, YOLOX model network structure, ScaledYOLOv4 Model network structure, YOLOv4 model network structure, YOLOv3 model network structure, YOLO-FaceV2 model network structure, TPH-YOLOv5 model network structure, SPD-YOLO model network structure, SlimNeck-YOLO model network structure, YOLOv5-Lite model network structure, PicoDet The model network structure, etc. are continuously updated...
🚀Includes various improved networks based on YOLOv5, YOLOv7, YOLOX, YOLOR, YOLOv3, YOLOv4, Scaled_YOLOv4, PPYOLO, PPYOLOE, PPYOLOEPlus, Transformer, YOLO-FaceV2, PicoDet, YOLOv5-Lite, TPH-YOLOv5, SPD-YOLO, etc.** Model configuration files for algorithmic models such as structures**
| Object Detection | Object Segmentation |
|---|---|
 |
 |
| Image Classification | Instance Segmentation |
|---|---|
 |
 |
| Object Segmentation | Object Tracking |
|---|---|
 |
 |
| Pose Estimation | Face Detection |
|---|---|
 |
 |
| Heat map 01 | Heat map 02 |
|---|---|
 |
 |
-
YOLOv5 https://github.com/ultralytics/yolov5/releases/tag/v6.1
-
Scaled_YOLO https://github.com/iscyy/yoloair/releases/tag/v1.0
About the code. Follow the design principle of YOLOv7.
The original version was created based on YOLOv7 and YOLOAir
Clone the version warehouse in the environment of Python>=3.7.0 and install requirements.txt, including PyTorch>=1.7.
$ git clone https://github.com/iscyy/yoloair2.git
$ cd yoloair2
$ pip install -r requirements.txt $ python train.py --cfg configs/yolov5/yolov5s.yamldetect.py runs inference on various data sources and saves the detection results to the runs/detect directory.
$ python detect.py --source 0
img.jpg
vid.mp4
path/
path/*.jpg Basically consistent with the YOLOv5 framework, you can refer toYOLOAir
- Train Custom Data 🚀 Recommended
- Tips for Best Training Results ☘️ Recommended
- Record experiments with Weights & Biases 🌟 NEW
- Roboflow: Datasets, Labels and Active Learning 🌟 New
- Multi-GPU training
- PyTorch Hub ⭐ New
- TFLite, ONNX, CoreML, TensorRT export 🚀
- Test Time Data Augmentation (TTA)
- Model Integration
- Model pruning/sparseness
- Hyperparameter Evolution
- Transfer learning with frozen layers ⭐ NEW
- Architecture Summary ⭐ NEW
In the future, we will continue to build and improve the YOLOAir ecosystem Perfectly integrate more YOLO series models, continue to combine different modules, and build more different network models Horizontal expansion and introduction of related technologies, etc.
@article{2022yoloair2,
title={{YOLOAir2}: Makes improvements easy again},
author={iscyy},
repo={github https://github.com/iscyy/yoloair2},
year={2022}
}Expand
-
The content of this site is only for sharing notes. If some content is infringing, please sending email.
-
If you have any question, please discuss with me by sending email.