A MXNet tiny face detector
Switch branches/tags
Nothing to show
Clone or download
chinakook Merge pull request #11 from sovrasov/master
Code refactoring for convenient use in other projects
Latest commit 2613ea0 Sep 6, 2018



This is an inference-only implementation for MXNet of tiny face. See the following references for more information:

"Finding Tiny Faces."
Peiyun Hu, Deva Ramanan
arXiv preprint arXiv:1612.04402, 2016.


Getting Started

  • Download the origin author's model from here.
  • Transform the origin Matconvnet model to MXNet model using matconvnet_hr101_to_mxnet.py.
  • To run:
    • simply run tiny_detection_mxnet.py
    • use from your code
from tiny_fd import TinyFacesDetector
import cv2 as cv

detector = TinyFacesDetector(model_root='./', prob_thresh=0.5, gpu_idx=0)
img = cv.imread('./selfie.jpg')
boxes = detector.detect(img)


You should install opencv-python 2.4 to get the nearly completely identical result to the origin implemetation.

Other implementations

cydonia999's Tensorflow implemetation