- detect_video.ipynb: detect videos from your own source
- evaluate_result.ipynb: evaluate performance of various detection models
- faster_rcnn.ipynb: use the Faster R-CNN model for detection
- ssd.ipynb: use the SSH model for detection
- training.ipynb: train the YOLOv3 model using custom dataset, annotation file and anchor file
https://www.anaconda.com/distribution/#download-section
conda env create -f environment.yml
conda activate tensorflow
jupyter notebook
- Open
detect_video.ipynb
in your browser - Change the
video_path
argument in the second cell to the path of your own input video - Change the
output_path
argument in the second cell to the desired output path - Run all cells