|
Introduction
this repository is the implementation of MTCNN in MXnet * core
: core routines for MTCNN training and testing. * tools
: utilities for training and testing * data
: Refer to Data Folder Structure
for dataset reference. Usually dataset contains images
and imglists
* model
: Folder to save training symbol and model * prepare_data
: scripts for generating training data for pnet, rnet and onet
Useful information
You're required to modify mxnet/src/regression_output-inl.h according to mxnet_diff.patch before using the code for training.
cache (created by imdb)
-- name + image set + gt_roidb
-- results (created by detection and evaluation)
mtcnn # contains images and anno for training mtcnn
-- images
---- 12 (images of size 12 x 12, used by pnet)
---- 24 (images of size 24 x 24, used by rnet)
---- 48 (images of size 48 x 48, used by onet)
-- imglists
---- train_12.txt
---- train_24.txt
---- train_48.txt
custom (datasets for testing)
-- images
-- imglists
---- image_set.txt
Results
License
MIT LICENSE
Reference
Kaipeng Zhang, Zhanpeng Zhang, Zhifeng Li, Yu Qiao , " Joint Face Detection and Alignment using Multi-task Cascaded Convolutional Networks," IEEE Signal Processing Letter