The octopus swims by actuating muscles embedded in the soft tissues.
SoftCon is an open source code that implements the work SoftCon: Simulation and Control of Soft-Bodied Animals with Biomimetic Actuators. With our framework, user can generate the swimming animation of under-water animals with deformable body simulator, biomimetic muscle pattern generator, and swimming controller based on deep reinforcement learning. This code is written in C++ and Python, based on Tensorflow and OpenAI Baselines.
Sehee Min, Jungdam Won, Seunghwan Lee, Jungnam Park, and Jehee Lee. 2019. SoftCon: Simulation and Control of Soft-Bodied Animals with Biomimetic Actuators. ACM Trans. Graph. 38, 6, 208. (SIGGRAPH Asia 2019)
Project page : http://mrl.snu.ac.kr/publications/ProjectSoftCon/SoftCon.html
Paper : http://mrl.snu.ac.kr/publications/ProjectSoftCon/SoftCon.pdf
Youtube : https://www.youtube.com/watch?v=I2ylkhPSkT4
Blog : http://mrl.snu.ac.kr/blog/ProjectSoftCon
We recommend users to install and run this framework on Ubuntu. We checked code works in Ubuntu 16.04 and 18.04.
sudo apt-get update
sudo apt-get install build-essential cmake-curses-gui git
sudo apt-get install libeigen3-dev freeglut3-dev libtinyxml-dev libpython3-dev python3-numpy libopenmpi-dev
- Download Boost 1.66 source code as zip file in https://www.boost.org/users/history/version_1_66_0.html.
- Unzip the file.
tar -xvf boost_1_66_0.tar.gz
- Compile and install the source code.
cd boost_1_66_0
sudo ./bootstrap.sh --with-python=python3
sudo ./b2 --with-python --with-filesystem --with-system install
sudo apt-get install python3-pip
sudo pip3 install virtualenv
virtualenv venv
source venv/bin/activate
pip install numpy
pip install scipy
pip install matplotlib
pip install tensorflow
pip install mpi4py
pip install OpenCV-Python
- OpenAI Gym
pip install gym
- OpenAI Baselines
sudo apt-get update && sudo apt-get install cmake zlib1g-dev
git clone https://github.com/openai/baselines.git
cd baselines
pip install -e .
git clone http://github.com/seiing/SoftCon
cd SoftCon
mkdir build
cd build
cmake ..
make -j8
./render/render
space
: play/pause.k
: play with default key.r
:reset the scene.
./render/render (network_name)
cd SoftCon/learn
mpirun -np 8 python3 -m run --type=train
- For bug reports/requests/suggestions, use github issues.
- For private inquiries, feel free to contact us at: [email protected]