- Python 3
- Pytorch 1.9+
- Numpy
- Opencv-python
- Scikit-image
- Scikit-learn
#conda
conda install pytorch torchvision torchaudio cudatoolkit=11.3 -c pytorch
#pip
pip install torch torchvision torchaudio
- Install latest Pytorch as default, for different verision please see pytorch.org
- Notice that Pytorch can be installed via conda or pip, if you suffering from network issues, please use pip with Tsinghua index url.
pip install torch torchvision torchaudio -i https://pypi.tuna.tsinghua.edu.cn/simple
- To validate Pytorch installation, type the statements below.
>>import torch
>>torch.cuda.is_availble()
#Get "True" when pytorch (GPU Version) successfully installed
# Download CACTI Code Library
git clone https://github.com/ucaswangls/cacti.git
#Install releative packages
cd cacti
pip install -r requirements.txt
#Compile and install CACTI
python setup.py develop
- If you suffering from network issues when running pip install -r requirements.txt command, please use the Tsinghua index url.
pip install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple/
Validate CACTI installation
python tools/test_cacti_repo.py
Text shown below when CACTI successfully installed
Welcome to the Video SCI repository!