Result video (SSD): https://www.youtube.com/watch?v=Vc1g27yJgsw
Result video (Mask-RCNN): https://www.youtube.com/watch?v=9N7jtUnCuz8
SSD keras 1
- ssd_download_essentials.ipynb: Run this notebook first to download SSD implemnetation, moviepy package and target video from github.
- SSD_car_detection.ipynb: Car detection using SSD is done in the first 6 cells.
SSD keras 2
- ssd_download_essentials_keras2.ipynb: Run this notebook first to download SSD implemnetation, moviepy package and target video from github.
- SSD_car_detection_keras2.ipynb: Car detection using SSD is done in the first 6 cells. This script will be in ./ssd_keras-master after running ssd_download_essentials_keras2.ipynb
Mask-RCNN
- Mask_RCNN_download_essentials.ipynb: This notebook git clone Mask-RCNN from matterport, download project video and pip install essential packages.
- visualize_car_detection.py: Utilities for visaulize Mask-RCNN result .Put this file in
./Mask_RCNN/
- Mask_RCNN_demo_car_detection.ipynb: Put this file in
./Mask_RCNN/
- Keras 1 or keras 2 (2 for Mask-RCNN)
- Tensorflow (1.3 for Mask-RCNN)
- Python 3