Prioritizing Aid from Above

By Jillian Anderson, Brian Gerspacher, Brie Hoffman

Fruit trees, such as coconuts and bananas, are an important source of food and income for many communities in the South Pacific. Unfortunately, these same communities tend to be at high risk for natural disasters such as hurricanes, tsunamis, and volcanic eruptions. Big data master's students Jillian Anderson, Brian Gerspacher, and Brie Hoffman participated in the 2018 Open AI Challenge hosted by WeRobotics with the ultimate goal of helping these communities receive aid more efficiently. Aid organizations typically collect and manually analyze aerial imagery to assess damage to affected areas. This analysis guides efficient distribution of aid, prioritizing communities whose food sources have been most affected by disaster. However, manual analysis of this data can take days to complete when time is of the essence. This delay hinders organizations from responding to crises in the most efficient manner possible. Anderson, Gerspacher, and Hoffman use computer vision and machine learning to automatically assess the damage caused by cyclones in the South Pacific by training a convolutional neural network to detect and count different kinds of trees present in aerial images. With its robust results, the trio's project stood out among the competition, winning the gold medal in the Open AI Challenge!

Project report | Project poster | Video