- B43B-03: USGS Flow Photo Explorer: A flexible AI/ML approach to transforming timelapse imagery into environmental insights
-
NOLA CC
Author(s):Generic 'disconnected' Message
Phillip Goodling, U.S. Geological Survey (First Author, Presenting Author)
Jennifer Fair, US Geological Survey Eastern Ecological Science Center
Amrita Gupta, Microsoft Corporation
Jeff Walker, Walker Environmental Research
Keegan Johnson, U.S. Geological Survey
Evan Gohring, US Geological Survey Colorado Water Science Center
Graham Sexstone, U.S. Geological Survey
Caitlin Marsteller, USGS Alaska Science Center
Hayley Olds, USGS Upper Midwest Water Science Center
Paul Reneau, USGS Upper Midwest Water Science Center
Sarah Schoen, USGS Alaska Science Center
This presentation describes an interactive data system and platform used to apply a deep learning artificial intelligence model to images collected at a fixed interval from inexpensive trail cameras, turning them into quantitative data. We demonstrate its use for several environmental monitoring applications including streamflow, algal blooms, ice and snow cover, and wildlife population density. Our system is novel in its ease-of-use and flexibility. We showcase the benefits of training multiple models on the same image timeseries to evaluate environmental processes. We solicit feedback on ways to make this system even more flexible and easy to use.
Scientific DisciplineNeighborhoodType
Enter Note
Go to previous page in this tab
Session


