@author: Luan Nguyen, Pengyu Chu, Daniel Shumaker
This repo is for the project in the class CSE 847 Machine Learning.
Structure:
codes/car_detection/
Final-Report/
Mid-Report/
Proposal/
[TOC]
The files you maybe need is on the Drive https://drive.google.com/drive/folders/1Eee7fgz57J7I4uVtdskzEfGFLv_vkppp?usp=sharing
git clone https://github.com/matterport/Mask_RCNN.git
cd Mask_RCNN
pip3 install -r requirements.txt
python3 setup.py installThe all required files are stored on the Google Drive
-
Add videos to aic19-track3-train-data/ or aic19-track3-train-data/
-
Add our model mask_rcnn_car_0030.h5 to model/; or train that from scratch.
-
git clone [Our Project] cd codes/car_detection/
# please ensure you are under the car_detection directory
python3 Mask_RCNN/car_detection_train.py train --dataset=dataset/ --weights=model/mask_rcnn_car_0030.h5 --logs=model/
# train from a pre-trained model COCO
# python3 Mask_RCNN/car_detection_train.py train --dataset=dataset/ --weights=model/mask_rcnn_coco.h5 --logs=model/
python3 infer_car_in_video.py
python3 sort_track.py
# output results into output_train/ or output_test/ in terms of your setup
# make sure there is at least one sequence file under the output_train/
python3 anomalies_detection.py
# please ensure you are under the car_detection directory
python3 Mask_RCNN/car_detection_train.py train --dataset=dataset/ --weights=model/mask_rcnn_car_0030.h5 --logs=model/
# train from a pre-trained model COCO
# python3 Mask_RCNN/car_detection_train.py train --dataset=dataset/ --weights=model/mask_rcnn_coco.h5 --logs=model/python3 infer_car_in_video.pypython3 sort_track.py
# output results into output_train/ or output_test/ in terms of your setup# make sure there is at least one sequence file under the output_train/
python3 anomalies_detection.py
# the output will be under the output_anomalies folder.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent