IS14D:
Moving Beyond Technology: How Quantitative Image-Based Methods Help Reveal Ocean Ecology II Posters
Moving Beyond Technology: How Quantitative Image-Based Methods Help Reveal Ocean Ecology II Posters
Session ID#: 85212
Session Description:
However, beyond the technological challenge of obtaining high quality images for scientific purposes, there is a need to scale up such measurements to obtain globally standardized observations, aided by advanced Machine Learning methods for image classification.
These developments will bring significant challenges for the coming years: 1) management, integration, and cross-calibration of the massive data flow originating from a network of different instruments; 2) automation of the classification of objects into taxonomic/morphological categories of scientific interest and import; 3) processing of data in near-real time, in particular for sensors embedded in autonomous platforms; 4) provision of collaborative pathways to visualize, annotate, quality control, and share the resulting data; and 5) integration of such data with existing multidisciplinary environmental sciences databases.
We invite presentations focusing on imaging and Machine Learning methods, but extending beyond pure technological challenges, in order to provide insights into ecological and biogeochemical processes in the ocean environment.
Co-Sponsor(s):
- ME - Marine Ecology and Biodiversity
- OB - Ocean Biology and Biogeochemistry
Index Terms:
1942 Machine learning [INFORMATICS]
4855 Phytoplankton [OCEANOGRAPHY: BIOLOGICAL]
4890 Zooplankton [OCEANOGRAPHY: BIOLOGICAL]
4894 Instruments, sensors, and techniques [OCEANOGRAPHY: BIOLOGICAL]