Skip to content

README.md

Latest commit

 

History

History
63 lines (50 loc) · 1.48 KB

README.md

File metadata and controls

63 lines (50 loc) · 1.48 KB

MEMFOF

How to Run (Pytorch)

  1. set up a virtual environment.
cd MEMFOF
git clone https://github.com/msu-video-group/memfof.git
cd memfof

# Create a new conda environment with Python 3.11
conda create -n memfof -y python=3.11

# Activate the created environment
conda activate memfof

# Install the required Python packages
pip3 install -r requirements.txt
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu128
  1. run the original pytorch model on test images.
python demo.py
  1. check pytorch model inference performance
cd ..
python ../gen_video2imgs.py
python infer.py
  • input size: 288 x 512, iters: 8
  • 249 iterations time: 41.1353 [sec]
  • Average FPS: 6.05 [fps]
  • Average inference time: 165.20 [msec]

How to Run (TensorRT)

  1. generate onnx file
python onnx_export.py
// a file '.onnx' will be generated in onnx directory.
  1. build tensorrt model and run
conda activate trte
python onnx2trt.py
// a file '.engine' will be generated in engine directory.
  • input size: 288 x 512, iters: 8
  • 249 iterations time: 17.6442 [sec]
  • Average FPS: 14.11 [fps]
  • Average inference time: 70.86 [msec]

Back