FaceBoxes: A CPU Real-time Face Detector with High Accuracy

A PyTorch Implementation of FaceBoxes: A CPU Real-time Face Detector with High Accuracy

Description

I train faceboxes with pytorch which approachs the official code,the final model can be downloaded in faceboxes.pytorch, the ap in AFW,PASCAL_FACE and FDDB as following:

AFW	PASCAL	FDDB
98.32	96.35	95.2

Requirement

• pytorch 0.3
• opencv
• numpy
• easydict

Prepare data

1. download WIDER face dataset
2. modify data/config.py
3. python prepare_wider_data.py

Train

python train.py --lr 0.001

Evalution

according to yourself dataset path,modify data/config.py,the evalution way is same with the official code

1. Evaluate on AFW.

python tools/afw_test.py

2. Evaluate on FDDB

python tools/fddb_test.py

3. Evaluate on PASCAL face

python tools/pascal_test.py

4. Test image

python demo.py

Result

References

• FaceBoxes
• FaceBoxes: A CPU Real-time Face Detector with High Accuracy