MTCNN Face Detection
This is a C++ version used Caffe
half model shrink the orignal onet by half quantities of parameters with faster forward computing and similar results on fddb you can try different thresholds and minsize value to perform best on your own datasets
-
video: ./MTCNN-Accelerate-Onet
image: ./MTCNN-Accelerate-Onet image imagefile
