This scripts will help us to understand the difference between dlib's HOG based and CNN based face detetction. I have provided some sample images and their output to understand.
- Clone the repository.
- Run hog_vs_cnn.py
python3 hog_vs_cnn.py -i/--image <image> -w/--weights <mmod weight> -o/--output <path to save ouput>
NOTE :
a. -w/--weight is optional if the mmod weights is present in the same directory else you have to provide the path of mmod weights in it.
b. -o/--output is also optional, if you want to save the result provide the path in it where you want to save.