This project aims to create a data mining pipeline to mine for fishing cast examples in creative commmons videos. I am using it to build up a model that I can use to gather analytics about my days on the water bass fishing.
- Data Mining Service.
- Model and Retraining Service.
- Human-in-the-Loop Verification Service.
- AWS S3 (currently with MinIO and boto3 for local dev env)
- RabbitMQ.
At first, I was not as concerned about code quality (as I just needed to make a prototype exist), but I am now working to iron out the existing features before moving on to further exploration of the model architecture. For example, I would love to expand to explore different models or even add more classes such as: side-arm cast, flipping cast, pitching cast, bite, and catch.