Validation of non-invasive measures for life stage in wild octopuses and AI-Assisted Behavioral Analysis
Grantee: Michaella Andrade
Institution: Federal University of ABC, Brazil
Grant amount: $10,000
Grant type: Seed grants
Focal species: Octopuses (Octopus insularis)
Conservation status: Not evaluated
Disciplines: Animal behavior
Research locations: Brazil
Project summary
This project aims to develop non-invasive methods to assess the life stage and welfare of wild octopuses. The project will develop an innovative AI approach that automatically learns to recognize complex behaviors, such as body patterns and ventilation rates, directly from pre-recorded videos in the field. In addition, the project will develop a methodology to automatically measure the life stage of octopuses in pre-recorded videos based on our database and non-invasive in situ measurements of the distance between the octopuses’ eyes and eyeballs. The main goal is to correlate these metrics with different behavioral contexts to determine whether they can be used as reliable indicators of the animal’s affective state. In the long term, the project aims to create an automated tool that accelerates and improves behavioral data analysis, enabling large-scale studies of the welfare of wild octopuses.
Why we funded this project
This project extends a previous WAI-funded project and uses innovative approaches to analyzing behavioral data for welfare assessment.