Awards & Nominations
Sparks 2.0 has received the following awards and nominations. Way to go!
Global Nominee
Ocean Heroes Design
High-Level Project Summary
Ocean Heroes (OH) is a game. Leveraging the power of humans, the game helps ML algorithms (YOLO) classify different types of ocean garbage. Players are rewarded with points for each successful labeling of garbage. Because many people play the game at the same time with the same set of images, a consensus can be reached about types of garbage on unclear images.The next stage is teaching a robot to sort the garbage in real time as people play the game. The OH robot will contain at least 6 different bags; one for cans, one for plastic bottles, one for packaging waste, etc. OH Robot will follow the garbage patches seen in the NOAA Garbage Patch Experiment until our oceans are completely clean!
Link to Project "Demo"
Link to Final Project
Detailed Project Description
Ocean Heroes is a game that allows players to reach crowd based consensus on the type of garbage remote sensing technology is capable of delivering to the players. It helps not only classify the garbage in pre-set categories but also allows players to introduce new categories to the list.
Playing this game allows AI/ML to further refine its identification algorithms and allows people play an active part in dealing with the issue of marine debris.
We hope to involve millions of people in cleaning up the oceans.
To design this game, we used Django Web Framework (Python), HTML, CSS, SQLite, YOLO(?)
Space Agency Data
The data used in this project was from the NOAA Garbage Patch Experiment. The dataset used for the ocean buoy visualization is the Global Drifter Database from the GDP Drifter Data Assembly Center, part of the NOAA Atlantic Oceanographic & Meteorological Laboratory.
We used this data to map the landscape of the game.
Hackathon Journey
Early on we discovered that none of us had worked with ML before. So we learned. We scoured the available resources for information on the technology available, datasets and imagery, how to upload a website to server, how to design a basic UX. Everything from scratch.
We had an additional issue of being all over the world in different timezones. So while some of us were sleeping, others kept working on their part of the project.
Setting up a to do list and assigning team members to each agreed item on the list helped us coordinate the work needed to be done.
We solved challenges together and helped each other on technical issues using collaboration tools available to us.
References
https://sos.noaa.gov/catalog/datasets/marine-debris-garbage-patch-experiment-drifters-and-model/
Tags
#AI, #Machine Learning, #ML, #MarineDebris, #Plastic, #Ocean, #Robot, #Clean
Global Judging
This project has been submitted for consideration during the Judging process.