Claire
Gorman
Hanly

Claire Gorman Hanly is an American computer scientist and environmental designer. Her research and practice focus on the implementation of deep learning-based computer vision methods in built and natural environments, with applications in regenerative agriculture, remote sensing of hydrology, and cultural landscape preservation.

Claire’s work ranges geographically across glaciers and caves, river deltas and grain supply-sheds, Arctic wilderness and subtropical cities. It ranges methodologically across scientific research, creative curation, and speculative tool-building. She has collaborated with technology companies, research labs, and the US National Park Service as well as her most recent role in co-curating the 19th International Architecture Exhibition at La Biennale di Venezia.

Claire is currently pursuing two Master’s degrees at MIT: an MCP in Environmental Planning, and an MS in Computer Science. Her Bachelor’s degree is in Computer Science and Architecture, from Yale University.

clairego@mit.edu
linkedin
instagram

Deep Learning on Ice

Implementation of an Attention U-Net for segmentation of glaciers in the Hindu-Kush region of the Himalayas, framed by discussion of image-based methods for documenting glacial extent over time. Based on original glacial segmentation paper by Baraka et al. (2020) and proposal of Attention U-Net by Oktay et al. (2018).

project link

image: Glaciated area of the Hindu Kush Himalayas in the disputed Jammu and Kashmir area of India and Pakistan, with clean-ice glaciers (training labels, from ICIMOD) shown in blue over Google Satellite data (2021).

Skills:
Deep Learning
Computer Vision
Remote Sensing

For MIT 6.S898 Deep Learning taught by Profs. Philip Isola and Stephanie Jegelka

2020-2022