From Realistic and Automated Fish Digital Twin to Data Ethics: Building Reproducible Pipelines and Dataset for Biological AI

" summary

Combines YOLOv8-based computer-vision segmentation of lab videos with the HAACHAMA CFD framework to produce dimensionless hydrodynamic digital twins of fish that replicate vortex dynamics and estimate pressure fields without PIV. Presents an open-source dataset and automated pipeline for reproducible biological-AI experiments, then outlines dataset design principles—transparent methodology, consistent tooling and annotations, ethical animal/data practices, durable open access, and licensing choices—to enable scalable, ethically grounded research.

AI-written summary · about

" related talks

• 25mDeep learning in deep nets: Helping fish farmers feed the worldBryton Shang · AWS re:Invent 2019 | AI & Machine Learning Summit
• 30mData-Centric AI Cookbook-let's prep that dataMarysia Winkels · PyData Global 2022
• 38mBig Data OceanographyJames Munroe · PyData London 2018
• 50mRestoring Coral Reefs at Scale: AI and the Next Generation of ConservationRich Williams · Decoding Communication in Nonhuman Species IV
• 28mDeeper Dive on Making Boats Fly with AI on RayNicolas Hohn · Ray Summit 2021
• 01h 00mData, visualization, and designing AIFernanda Viegas, Martin Wattenberg · rstudio::conf 2019
• 28mTrusted AI – Building Reproducible, Unbiased and...Romeo Kienzler · PyData Warsaw 2019
• 24mAI in maritime - Autonomous shipsMarina Ivasic Kos · DSC Europe 24