Vision Object Detection

[LianAustin]

Create a functional template for object detection. Ensure there is a reliable way to train a model.

Create a reliable vision model for the centerstage object. ----------------------------------------------------------------goals-----------------------------------------------------------------------

• [ ] be able to train an image detection model without using the ftc model builder
• [x] be able to convert the saved model into a tflite format
• [ ] be able to run the tflite model on the robot ---------------------------------------------------------------solutions----------------------------------------------------------------------
• [ ] train a tf1 model
• [x] add metadata to the tf2 model
• [x] figure out why there are 8 outputs to the current model
• [ ] add metadata to the tf2 saved model format
• [ ] convert the tf2 model into a tf1 model