Frameworks: PyTorch, OpenCV, Face_Detection
Navigate to source code directory and install the environment either using conda or requirements.txt followed by pip install command.
conda env create --file conda-gpu.yml
pip3 install -r requirements.txt
After installation or activating the enviroment, execute the following command to run over live input stream:
python3 object_tracker.py --video 0 --output $path_to_output_video
To run over existing clip, run the following command:
python3 object_tracker.py --video $path_to_input_video --output $path_to_output_video
Finally, .csv and .txt files are generated within the source_code directory after completion of the above command.
Link to google drive for model generated output videos, text and csv files are here.
