README.md

ML-CV

Machine Learning-Computer Vision

What is ML-CV?

ML-CV is our 2021 Vision system which combines the benefits of a traditional OpenCV system with that of a Machine Learning model. We use traditioanl OpenCV filtering and masking tecniques alongside a TensorFlow Lite Object Detection model to accurately detect and pinpoint the location of a vision target or gamepiece.

How does ML-CV work?

or in meme form...

Train custom TFLite model

• COCO dataset: Roboflow
• TFRecord dataset: Roboflow
• Train SSDMobileNet v2 model: Google Colab NB
• Convert to TFLite: Google Colab NB