Restoring Coral Reefs at Scale: AI and the Next Generation of Conservation
summary
Coral Gardeners is running deployable field tech: an iPhone data app called Reef app, a solar-powered, connected reef cam with stereo cameras and hydrophone, and a prototype 3D photogrammetry pipeline that combines SegFormer benthic segmentation with point‑cloud voting. Labeled assets include ~2–3k donor‑colony photos and ~2k fish images; YOLOv8 gets ~85–90% precision/recall, unsupervised audio embeddings produce ~20 sound clusters and ~1.5× more loud events on restored vs control sites. Failures include a clunky 3D UI and fragile tethered hardware; next steps focus on scalable battery cameras and DNA integration.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 1:09Coral Gardeners mission
- 5:19Coral gardening process
- 9:13Reef app replaces manual data
- 10:59Underwater camera system
- 13:26AI for ecosystem restoration
- 16:45Photogrammetry for reef monitoring
- 23:58Bioacoustic analysis of reefs
- 32:21AI for fish population
- 41:16AI conservation challenges
- 44:28Summary
related talks
05mA journey to protect the Great Barrier Reef using Machine Learning
29mHow a Passion for Oceans Can Utilize Synergies of Technology
28mConservation & Tech
29mBioacoustics and Machine Learning as Key Tools in Conservation
47mProduction ready EdgeAI with Eclipse ioFog and Google Coral
15mThe Internet of Wild Things: Acoustics, AI and Conservation
25mDeep learning in deep nets: Helping fish farmers feed the world
16mStartups can save our oceans