Hello welcome to my face detection
its a python model which can detect faces in images with 1 face at most
Ubuntu " It's only tested on Ubuntu, so it may not work on Windows "
GPU : Any GPU that is works with PyTorch
vram : 4gb if you want to train ' we train it in rtx 2070 '
numpy : https://numpy.org/
PyTorch : https://pytorch.org/
torchvision : https://pypi.org/project/torchvision/
openCV : https://pypi.org/project/opencv-python/
- Download my pretrained model : https://drive.google.com/file/d/1sq1eZhRyQya1G8Lsdb5aqfVwEg5FhMH0/view?usp=sharing
OR
-
Download the faces DataSet and put it in project directory: http://vis-www.cs.umass.edu/fddb/
-
Download the no faces TrainSet and TestSet :
https://drive.google.com/drive/folders/1lrSYMFTtP7k_orCdkes-Vl5dgtGPaJUy?usp=sharing
https://drive.google.com/drive/folders/1vCMbBfaU6fLOsTNh6qrRNp5kqNKNW_nV?usp=sharing
-
Run the train.py file :
$ python3 train.py
-
Test an image
$ python3 test_image.py [your image path]
-
Test a video
$ python3 test_video.py [your video path]