CoralGuard

• 

  The flask application (python w/ html)

Inspiration

Due to climate change, coral algae (which give corals their color) is stressed by the heat and causes coral bleaching. Corals are vital for ocean communities, so by identifying the problem early, there can be early prevention.

What it does

Detects signs of coral bleaching from images.

How it's built it

• ML model with kaggle dataset & Google Collab (trained with pytorch)
• Blood, sweat, and tears (Html, Javascript, css, and flask)

Challenges ran into

Training the ML model was particulary challenging since I trained the model on images, so it took a long time for the model to iterate over the dataset. Additionally, the model would time out occasionally so I had to implement checkpoints for the ML model

What's next for CoralGuard

Implementation in an actual underwater camera

Extra Credits: Special thanks to Isabella for collaborating with me to train the AI model (Helped with dataset uploading & creating a hyperparameter search for dropout and learning rates).

Built With

• flask
• python
• pytorch