Neural‑Net Powered Offline Android App for Real‑Time Fish Species, Health, and Volume Estimation

The project develops an on‑device AI‑driven Android application that enables fishermen, inspectors, and seafood buyers to capture images of fish catches and instantly obtain species identification, freshness/health assessment, and quantitative volume/weight estimates. The solution operates offline, uses lightweight neural models (YOLOv8‑Tiny, EfficientNet‑Lite, custom CNN), leverages ARCore for scale, stores geotagged records locally, and optionally syncs with governmental or market cloud platforms for traceability and sustainable fisheries management.

This is a Complete Project Template

Sign in to view the full implementation details, methodology, and resources.