Home

Awesome

mtcnn-caffe

Joint Face Detection and Alignment using Multi-task Cascaded Convolutional Neural Networks.

This project provide you a method to update multi-task-loss for multi-input source.

result

Requirement

  1. Ubuntu 14.04 or 16.04
  2. caffe && pycaffe: https://github.com/BVLC/caffe
  3. cPickle && cv2 && numpy

Train Data

The training data generate process can refer to Seanlinx/mtcnn

Sample almost similar to Seanlinx's can be found in prepare_data

Strongly suggest readers generate training data themselves. The sample training data of 12net and 24net is available( Too big for Baidu Drive to upload 48net) by sending Email

Net

The main idea is block backward propagation for different task

12net 12net 24net 24net 48net 48net

Questions

The Q&A bellow can solve most of your problem.

Q1: What data base do you use?<br/> A1: Similar to official paper, Wider Face for detection and CelebA for alignment.

Q2: What is "12(24/48)net-only-cls.caffemodel" file for?<br/> A2: Provide a initial weigh to train. Since caffe's initial weigh is random, a bad initial weigh may take a long ran to converge even might overfit before that.

Q3: Why preprocess images by minus 128?<br/> A3: Separating data from (0,+) to (-,+), can make converge faster and more accurate. Refer to Batch Normalization Accelerating Deep Network Training by Reducing Internal Covariate Shift

Q4: Do you implement OHEM(Online-Hard-Example-Mining)?<br/> A4: No. OHEM is used when training data is not that much big. Refer to faster-rcnn's writer RBG's paper

Q5: Ratio positive/negative samples for 12net?<br/> A5: This caffemodel used neg:pos=3:1. Because 12net's function is to eliminate negative answers, similar to exclusive method, we should learn more about negative elininate the wrong answer.

Q6: Why your stride is different to official?<br/> A6: If you input a (X,X) image, the output Y = (X-11)/2. Every point on output represent a ROI on input. The ROI's left side moving range = (0, X-12) on input, and (0, Y-1) on output. So that stride = (X-12)/(Y-1) ≈≈ 2 in this net.

Q7: What is roi(cls/pts).imdb used for?<br/> A7: Use imdb can feed training data into training net faster. Instead of random search data from the hard-disk, reading data from a large file once to memory will save you a lot of time. imdb was created by python module-cPickle.

Q8: What is tools_matrix.py different from tools.py?<br/> A8: Matrix version use linear matrix to make calculation faster(160ms on FDDB). If you are green hand in this area, read Non-Matrix version to understand each process.

Q9: I want your training data for emergency use. How to use them? How to train? ...<br/> A9: ???

Current Status

CongWeilin updated in 2017/3/5

Update tools_matrix.py to make calculate faster, about 160ms/image.