Skip to content

Latest commit

 

History

History
60 lines (30 loc) · 1.52 KB

README.md

File metadata and controls

60 lines (30 loc) · 1.52 KB
Raw

RigMe

demo

Python Version

3.7.2

Possible libraries needed to install

brew install spatialindex

pip install google-colab torch torchvision scikit-image trimesh torch-cluster open3d torch_geometric torch_sparse torch_scatter torch_scatter

pip install google-colab torch torchvision scikit-image trimesh torch-cluster open3d torch_geometric torch_sparse torch_scatter gluoncv pycocotools mxnet tensorboard argparse rtree

rtree:

pip install Rtree

For Mac: sudo pip3 install osmnx

#Github Downloads in specificed paths follow the setup procedure for each github repo, mainly in the image background remove tool and lightweight human pose

./image_background_remove_tool: https://github.com/OPHoperHPO/image-background-remove-tool

./lightweight_human_pose_estimation: https://github.com/Daniil-Osokin/lightweight-human-pose-estimation-3d-demo.pytorch

./pifuhd: https://github.com/facebookresearch/pifuhd

./RigNet_Master: https://github.com/zhan-xu/RigNet

Usage

First run the flaks application:

python convert_image.py

Then use this command to send a picture to convert into a 3d model:

curl -X POST http://0.0.0.0:80/api/convert_picture --form file=@./sample_images/test_v4.jpg --form resolution=78

Setting up Blender

download:

https://www.blender.org/download/

execute: echo "alias blender=###PATH_TO_BLENDER.APP/Contents/MacOS/blender" >> ~/.bash_profile

#####IMPORTANT CODE######## Xvfb :19 -screen 0 1024x768x16 & export DISPLAY=:19