Data Dynamos GAT -13

Inspiration

Extinction of the rare species

What it does

To understand the intricate dynamics of marine debris, our model prototype incorporates information from a variety of data sources, such as wind patterns, satellite photos, and data on ocean currents. These datasets allow us to forecast the amounts and locations of marine debris in different marine regions. Furthermore, because the initiative aims to use technology to improve society and the environment, it is in line with the ideals of "Tech for Good." We want to address one of the most important environmental concerns of our day by increasing awareness, encouraging appropriate trash disposal, and providing data-driven solutions. We have made great progress towards a cleaner and more sustainable marine environment with our predictive model prototype.

How we built it

We constructed this model by utilizing frameworks and libraries, including TensorFlow, PyTorch, and OpenCV. We sourced input data from labeled debris images available on Kaggle and conducted our work using Google Colab and Jupyter Notebook.

Challenges we ran into

At first, the plan was to obtain images for a specific region from Google Earth Engine through their API. However, enabling this option in the API came with significant costs, so we made a spontaneous decision to utilize images with debris instead.

Accomplishments that we're proud of

The completed model with appropriately labeled classification images.

What we learned

Through this datathon, we gained valuable experience in teamwork, problem-solving, time management, networking with both like-minded individuals and industry professionals, boosting our confidence, and learning from our mistakes.