Article Directory
foreword
Recently, I want to use YOLOv5Face to do a license plate detection task. Although the license plate detection is a bad street now, I still want to try it. YOLOV5Face is also a face detection algorithm that I think is very good, so I changed it on it. However, due to limited resources, I did not wait for the training to complete, so the pre-trained model will not be provided.
1. Preparation
1. Code path
The code address is: Code
Just download the code directly.
3. Environment installation
Environment installation according to requirements.txt
pip install -r requirements.txt
2. Data set preparation
The data set layout format is as follows:
Among them, the images are all our pictures, and the labels are the label files we generated, as follows: To
generate the label files, you can refer to my blog: CCPD data set processing
Two, training
After setting the parameters in trian.py, just run it directly, or execute the following code:
python train.py
3. Test
Set the parameters in detect_face.py, and then run it directly.
python detect_face.py
4. Results display
Here I put nearly 6w pictures for training and testing. Personally, I don’t feel the need to use all the data sets. After training for 70epoch, the training indicators such as P value, R value, and MAP are as follows:
the measured results are as follows:
Summarize
The above is the entire content of this article. If you have any questions, please communicate in the comment area.