Competition
IEEE YESIST 2026 — Foursight Labs
AI-based fish freshness decision support system — turning subjective visual inspection into objective, transparent food-safety screening.
FIELD DOCUMENTATIONPHOTO · COMPETITION
01 — Overview
Foursight Labs is an AI decision support system that classifies fish freshness objectively. It addresses the inconsistency of human-only visual inspection in food markets by replacing subjective judgment with data-driven screening.
The project aligns with SDG 3 (Good Health and Well-being) and SDG 12 (Responsible Consumption and Production), supporting safer food ecosystems in traditional market environments.
02 — Key Contributions
- Core model: MobileNetV2 CNN — lightweight architecture for real-time mobile inference without heavy server infrastructure.
- Explainable AI: Grad-CAM heatmaps highlight degradation areas (cloudy eyes, pale gills) to build user trust beyond black-box predictions.
- Outputs: binary Fresh / Not Fresh classification, confidence score, and visual heatmap for validation.
- Designed for deployment in traditional markets where portable, on-device screening matters most.