Namespace:    coralnet

PI: David Kriegman
Institution: University of California, San Diego
Project description:

Global and local stressors have caused a rapid decline of coral reefs across the world. To monitor the changes and take appropriate action large spatio-temporal surveys are needed. Data collection speeds are typically sufficient to meet this need but the subsequent image analysis remains slow as manual inspection of each photo is required. This creates a ‘manual annotation bottleneck. CoralNet reduces this bottleneck by allowing modern computer vision algorithms based on deep learning to be deployed alongside human experts. Deployed as a web site and service, is a platform for collaboration and sharing of data.

Software: Python, CUDA, Pytorch, Caffe

Back to namespaces