Skip to content

pzhaojing/Mask-RCNN-TF_detection-human_segment-body_keypoint-regression

 
 

Repository files navigation

Detecting human body parts and Building Skeleton Models using Deep Convolutional Neural Networks

This repository contains an extension to the amazing work done by CharlesShang This is a model of a neural network that is able to do object detection, classification and segmentation. The ideea is to detect all persons in an image and segment their body parts. The next step is to add keypoint regression. This is a real-time model running at 200ms/frame on a Titan X GPU. A practical usage for this kind of model would be to a fashion application that looks at the information of your body and proposes the look of different clothes in order for the user to visualize their appearance. Here are some results from the training set.

demo demo demo demo demo demo demo

The following videos contain the model that does body segmentation (the rest of the parts are not drawn). The model output has been modified to have 2 classes for detection (human, non-human) and 7 segmentation classes (full body, head, torso, right hand, left hand, right leg, left leg).

IMAGE ALT TEXT HERE IMAGE ALT TEXT HERE

The following videos shows the results of the same model, but this time are drawn the body parts.

IMAGE ALT TEXT HERE

Installation

git clone https://github.com/Iftimie/MaskRCNN_body.git
cd MaskRCNN_body
mkdir data/coco
mkdir data/coco/records
cd data/coco/records/

wget https://www.dropbox.com/s/43ihvomchvwtpns/checkpoint
wget https://www.dropbox.com/s/v6084wee6pjlfk4/coco_resnet50_model.ckpt-248000.data-00000-of-00001
wget https://www.dropbox.com/s/0gqxnbsjzpuz0tz/coco_resnet50_model.ckpt-248000.index
wget https://www.dropbox.com/s/3uildv0wlh79oad/coco_resnet50_model.ckpt-248000.meta
#modify the checkpoint file with your path
cd ../../..
git checkout test
#modify line 180 in train/test.py with your ip address
#modify line 36 in train/client.py with the respective ip address

#in one terminal
CUDA_VISIBLE_DEVICES=0 python train/test.py

#in another terminal. Make sure to have a webcam connected
python train/client.py

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 93.7%
  • C++ 2.4%
  • C 1.9%
  • Cuda 0.9%
  • Shell 0.7%
  • MATLAB 0.4%