Our projects


AI for fish in New England

The databranch built a team of fisheries and machine learning experts to create the N+1 fish, N+2 fish challenge, where competitors developed algorithms to automatically count, measure, and identify fish species in video from actual fishing vessels. The open source results will help make it easier and faster to monitor fishing in New England, providing better data to sustain future fish populations.

Project partners included CVision AI, DrivenData, and The Nature Conservancy and funded by a grant from the National Fish and Wildlife Foundation.


Improving U.S. Fisheries Data Systems

As part of a 2016 expert panel, databranch founder Kate Wing helped develop recommendations to increase the accuracy, speed, and usability of U.S. fisheries data. The databranch supports ongoing projects to implement those recommendations.


Citizen Science & Policy

In the summer of 2017, the databranch designed and facilitated a discussion at Berkeley Law with legal and data experts to explore the ways participatory, open, and citizen science can be used to make policy decisions.