zeerafle/cctv-inference

Accident Detection Inference API

How to run

1. Create a file named .flaskenv in the root directory, put the following content in it:

   FLASK_APP=cctv.py
   FLASK_ENV=development
   FLASK_DEBUG=1
   AWS_ACCESS_KEY_ID=<your AWS access key>
   AWS_SECRET_ACCESS_KEY=<your AWS secret access key>
   BUCKET_NAME=<your bucket name>
   

2. Install requirements (or create a virtual environment beforehand), e.g.:

   pip install -r requirements.txt

3. Get the model here
4. Install docker or podman
5. Run redis server via docker/podman:

   docker run -d -p 6379:6379 redis

   For podman, follow instructions here
6. Run redis worker:

   rq worker saving-tasks

7. Open another terminal, run the inference server:

   flask run

8. Go to your inference server's URL, e.g. http://localhost:5000, follow the instruction there to test the API.

The predicted frame is stored in AWS S3 bucket.

Example

TODO

• Store the predicted frames
• Write test
• Connect to cloud bucket storage
• Use redis worker to queue the upload to cloud storage process
• Actually deploy it