Setting up the environment

STEP 1 : Create a new conda virtual env with python 3.10 and activate it

conda create --name openmmlab python=3.10.15 -y
conda activate openmmlab

STEP 2: Install PyTorch 2.0.1 in the virtual env

OSX

conda install pytorch==2.0.1 torchvision==0.15.2 -c pytorch

Linux and Windows:

CUDA 11.8

conda install pytorch==2.0.1 torchvision==0.15.2 pytorch-cuda=11.8 -c pytorch -c nvidia

CPU Only

conda install pytorch==2.0.1 torchvision==0.15.2 cpuonly -c pytorch

STEP 3: Installing other required libraries (Install the exact versions below)

pip install openmim
mim install mmengine==0.8.2
mim install "mmcv>=2.0.0,<2.1.0"
mim install mmdet==3.1.0
pip install roboflow
pip install yapf==0.40.1

STEP 4: Some libraries may require a previous version, install the specific versions from the requirements.txt

Starting the state detection and move suggestion

STEP 1: Start a shareable video stream from your smartphone using an app like IP Webcam

Make sure the smartphone and the device running the ./game/gui.py script are on the same network

STEP 2: Set VIDEO_URL='{IP Address of the video stream}/video' in ./game/gui.py

STEP 3: Run ./game/gui.py, it will start the GUI for the Game

STEP 4: Click on the Detect Board State Button in the GUI to detect the game state from the stream