COTS-Detection-using-Efficient-Det-D1-in-TFOD-2.0-using-Tensorflow-great-brief-barrier-reef-dataset

This project is still work on progress

I had tried this project on EfficientDet D1 in TFOD 2.0 api. I had followed many notebooks, discussion, and blogs. I had used almost 7 types of Bounding box augmentation here. The notebooks I had followed for this:

1. https://www.kaggle.com/code/khanhlvg/cots-detection-w-tensorflow-object-detection-api
   2)https://www.kaggle.com/code/ravishah1/cots-faster-rcnn-training-w-tf-2-0-od-api-0-474/notebook
   

2. https://www.kaggle.com/code/datastrophy/cots-tfod2-api-setup-faster-rcnn
   

3. https://www.kaggle.com/code/khanhlvg/inference-using-efficientdet-d0-model-tensorflow
   

4. https://www.kaggle.com/code/andradaolteanu/greatbarrierreef-full-guide-to-bboxaugmentation
   

Discussion helped me lot

1. https://www.kaggle.com/competitions/tensorflow-great-barrier-reef/discussion/296976
   

Blog I followed:

1. https://tensorflow-object-detection-api-tutorial.readthedocs.io/en/latest/
   

2. https://stackoverflow.com/questions/44906317/what-are-possible-values-for-data-augmentation-options-in-the-tensorflow-object
   

I had already download the dataset to my google drive and mount the drive here

Output: