# CoachAI Machine Setup

## Install Package

### Environments

Ubuntu 20.04
python3 == 3.8.10

### install

sudo apt-get upgrade
sudo apt install -y python3-pip
pip3 install PyQt5==5.15.4
pip3 install opencv-python==4.2.0.34
pip3 install paho-mqtt
pip3 install playsound
pip3 install imutils
pip3 install pandas
pip3 install PyOpenGL
pip3 install joblib
pip3 install scipy
pip3 install tqdm
pip3 install numba
pip3 install filterpy
pip3 install yacs
pip3 install smplx
pip3 install trimesh
pip3 install pyrender
pip3 install scikit-image
pip3 install h5py
pip3 install pytube
pip3 install scikit-learn
sudo apt-get install -y freeglut3
sudo apt-get install -y freeglut3-dev
sudo apt install -y mosquitto
sudo apt install -y mosquitto-clients
sudo apt install -y ffmpeg
sudo apt-get install -y gstreamer1.0-plugins-good gstreamer1.0-plugins-bad gstreamer1.0-plugins-ugly gstreamer1.0-libav

## Install Nvidia GPU driver
[How To Install Nvidia Drivers On Ubuntu 20.04 / Ubuntu 18.04](https://www.itzgeek.com/post/how-to-install-nvidia-drivers-on-ubuntu-20-04-ubuntu-18-04.html)

### Install Nvidia driver from Ubuntu repository

sudo apt update
sudo ubuntu-drivers devices
sudo ubuntu-drivers installapt
sudo reboot

### Graphical Install
Goto Application `Additional Drivers`

and choose the driver you want to install

### Check if the driver is installed properly
Type `nvidia-smi` in the command line
If succeed, it should look like below

If not, checkout this article
> [ubuntu环境下，系统无法与NVIDIA通信的解决方法](https://wangpei.ink/2019/01/19/NVIDIA-SMI-has-failed-because-it-couldn't-communicate-with-the-NVIDIA-driver%E7%9A%84%E8%A7%A3%E5%86%B3%E6%96%B9%E6%B3%95/)

## Install CUDA Toolkit
First checkout the version you should install

[CUDA Toolkit 11.5 Downloads](https://developer.nvidia.com/cuda-downloads)
[CUDA Toolkit 11.3 Downloads](https://developer.nvidia.com/cuda-11.3.0-download-archive)

## Install Pytorch-GPU
https://pytorch.org/get-started/locally/

## Tcam SDK (0.14)
[Tcam Documentation](https://www.theimagingsource.com/documentation/tiscamera/tutorial.html#setup)
[Github link](https://github.com/TheImagingSource/tiscamera)

### Cloning

git clone https://github.com/TheImagingSource/tiscamera.git
cd tiscamera
mkdir build

### Dependencies

./scripts/dependency-manager install

### Compilation

make -j

### Building tiscamera

cd build
cmake -DBUILD_ARAVIS=ON ..
make
sudo make install

## VIBE
[Github link](https://github.com/mkocabas/VIBE)

### cloning

git clone https://github.com/mkocabas/VIBE.git

### install the package

sudo apt install python3.8-venv
python3.8 -m venv vibe-env
pip3 install git+https://github.com/giacaglia/pytube.git --upgrade
pip3 install tqdm
pip3 install yacs
pip3 install h5py
pip3 install numpy
pip3 install scipy
pip3 install numba
pip3 install smplx
pip3 install gdown
pip3 install PyYAML
pip3 install joblib
pip3 install pillow
pip3 install trimesh
pip3 install pyrender
pip3 install progress
pip3 install filterpy
pip3 install matplotlib
pip3 install tensorflow
pip3 install tensorboard
pip3 install torchvision
pip3 install scikit-image
pip3 install scikit-video
pip3 install opencv-python
pip3 install llvmlite
pip3 install git+https://github.com/mattloper/chumpy.git
pip3 install git+https://github.com/mkocabas/yolov3-pytorch.git
pip3 install git+https://github.com/mkocabas/multi-person-tracker.git

### download data

cd VIBE
source scripts/prepare_data.sh
cp -r data/ ~/