- Ubuntu 22.04 LTS
- Python 3.10
- PyTorch 2.0.1
- CUDA 11.8
a. Upgrade version of g++/gcc to 5+ for building c++ files
b. Create a conda virtual environment and activate it.
conda create -n ChangShuoRadioRecognition python=3.10 -y
conda activate ChangShuoRadioRecognition
c. Install PyTorch and torchvision.
conda install pytorch torchvision torchaudio pytorch-cuda=11.8 -c pytorch -c nvidia
d. Install mmdet for radio detection and radio recognition.
pip install -U openmim
mim install mmengine
mim install mmdet
e. Clone the ChangShuoRadioRecognition repository.
git clone https://github.com/ChangShuoRadioRecognition.git
cd ChangShuoRadioRecognition
f. Install build requirements and then install ChangShuoRadioRecognition.
mim install -e . # or "python setup_backup.py develop"
g. Install "Times New Roman" for matplotlib, following the link
sudo apt-get update
sudo apt-get install ttf-mscorefonts-installer
rm -rf ~/.cache/matplotlib
h. Install tsnecuda following the link.