# Instructions
MTCNN is a face detection model
The Chinese introduction can be found [here](https://blog.csdn.net/qq_32478489/article/details/106193921)
[paper](https://kpzhang93.github.io/MTCNN_face_detection_alignment/paper/spl.pdf)
Referring to [kuaikuaikim/DFace](https://github.com/kuaikuaikim/DFace) and [Sierkinhane/mtcnn-pytorch](https://github.com/Sierkinhane/mtcnn-pytorch), I fixed some bugs that would appear in the training, added the table of learning rate, and optimized the training parameters.Testing on the WiderFace validation set, I found it performance better than the model weight they had originally provided.
#### Test image
# WiderFace Val Performance in MTCNN
MTCNN-original is the test result of the original weight parameter
MTCNN-trained is the test result of my training
| Style | easy | medium | hard |
|:-|:-:|:-:|:-:|
| MTCNN-original | 65.3% | 65.1% | 40.3% |
| MTCNN-trained | **71.4%** | **70.4%** | **43.2%** |
# Installation
1.pytorch
2.opencv
# test
### a single picture
Modify the image path in the program
```Shell
python detect.py
```
# Train the model
download [widerface](http://shuoyang1213.me/WIDERFACE/)
Organise the dataset directory as follows:
```Shell
./data_set/face_detection/
WIDER_train/
images/
WIDER_val/
images/
```
I have made the label file
```Shell
./anno_store/anno_train.txt
```
If you want to make your own labels
The reference program
./anno_store/tool/change.py
### train pnet
**prepare pnet data**
```Shell
python mtcnn/data_preprocessing/gen_Pnet_train_data.py
python mtcnn/data_preprocessing/assemble_pnet_imglist.py
```
**train pnet**
```Shell
python mtcnn/train_net/train_p_net.py
```
**prepare rnet data**
```Shell
python mtcnn/data_preprocessing/gen_Rnet_train_data.py
python mtcnn/data_preprocessing/assemble_rnet_imglist.py
```
**train rnet**
```Shell
python mtcnn/train_net/train_r_net.py
```
**prepare onet data**
```Shell
python mtcnn/data_preprocessing/gen_Onet_train_data.py
python mtcnn/data_preprocessing/assemble_onet_imglist.py
```
**prepare onet**
```Shell
python mtcnn/train_net/train_o_net.py
```
# Evaluation widerface val
```Shell
python wildface_test.py
```
# References
[kuaikuaikim/DFace](https://github.com/kuaikuaikim/DFace)
[Sierkinhane/mtcnn-pytorch](https://github.com/Sierkinhane/mtcnn-pytorch)
MTCNN-Pytorch:人脸检测模型 （104个子文件） 
val-medium.JPG 187KB val-hard.JPG 182KB val-easy.JPG 175KB 23.jpg 140KB 1.jpg 126KB result.jpg 119KB 303.jpg 63KB 304.jpg 21KB
wider_face_train.mat 1.48MB wider_face_val.mat 388KB wider_face_test.mat 90KB README.md 3KB
o.png 215KB r.png 185KB p.png 178KB onet_epoch_train.pt 880KB onet_epoch.pt 879KB rnet_epoch_train.pt 238KB rnet_epoch.pt 237KB pnet_epoch_train.pt 29KB pnet_epoch.pt 29KB detect.py 28KB train.py 11KB models.py 9KB resnet_inception_v2.py 9KB gen_Rnet_train_data.py 8KB gen_Onet_train_data.py 7KB gen_Pnet_train_data.py 6KB imagedb.py 5KB vision.py 5KB gen_landmark_48.py 5KB image_reader.py 4KB utils.py 4KB train_r_net.py 3KB train_p_net.py 3KB wildface_test.py 2KB utils.py 2KB image_tools.py 1KB nms.py 1KB train_o_net.py 1KB assemble.py 1KB change.py 1KB config.py 1KB mtcnn_test.py 1001B assemble_pnet_imglist.py 689B assemble_rnet_imglist.py 689B assemble_onet_imglist.py 688B __init__.py 0B __init__.py 0B __init__.py 0B __init__.py 0B __init__.py 0B detect.cpython-35.pyc 15KB detect.cpython-36.pyc 14KB models.cpython-35.pyc 8KB models.cpython-36.pyc 7KB train.cpython-35.pyc 7KB image_reader.cpython-35.pyc 6KB BBox_utils.cpython-35.pyc 6KB train.cpython-36.pyc 5KB image_reader.cpython-36.pyc 5KB imagedb.cpython-35.pyc 5KB imagedb.cpython-36.pyc 4KB vision.cpython-35.pyc 3KB vision.cpython-36.pyc 3KB utils.cpython-35.pyc 3KB utils.cpython-36.pyc 3KB landmark_utils.cpython-35.pyc 3KB utils.cpython-35.pyc 2KB utils.cpython-36.pyc 2KB image_tools.cpython-35.pyc 1KB image_tools.cpython-36.pyc 1KB config.cpython-35.pyc 1KB config.cpython-36.pyc 1KB assemble.cpython-35.pyc 1KB assemble.cpython-36.pyc 885B __init__.cpython-35.pyc 165B __init__.cpython-35.pyc 160B __init__.cpython-35.pyc 155B __init__.cpython-36.pyc 154B __init__.cpython-36.pyc 149B __init__.cpython-36.pyc 139B
wider_face_train_bbx_gt.txt 4.72MB anno_train.txt 3.06MB imglist_anno_48.txt 2.9MB wider_face_val_bbx_gt.txt 1.17MB wider_face_test_filelist.txt 857KB readme.txt 558B tips.txt 75B tip.txt 40B tips.txt 25B tips.txt 25B tips.txt 25B tips.txt 25B tips.txt 23B landmark_48.txt 0B neg_48.txt 0B part_24.txt 0B imglist_anno_12.txt 0B imglist_anno_24.txt 0B共 104 条
- 1
- 2
评论0
最新资源