A non-invasive, vision-based system for estimating the physical measurements of Japanese quails (Coturnix japonica) from images — developed as a research project under Iloilo Science and Technology University (ISAT-U). Given a photo or video frame of a quail, the system predicts weight, height, width, and length using a CNN multi-output regression model, eliminating the need for stressful manual handling.

Built with Flutter for cross-platform mobile and web deployment, backed by a FastAPI inference server running an EfficientNet-B0 regression model. On-device inference is supported via a quantized MobileNetV3-Small TFLite model, enabling offline use in low-connectivity farm environments across the Philippines.

• InstitutionIloilo Science and Technology University (ISAT-U)
• PlatformMobile & Web Application (Flutter)
• StackFlutter, FastAPI, TensorFlow / Keras, EfficientNet-B0, MobileNetV3-Small, TFLite, Python
• ML TaskCNN Multi-Output Regression (Weight, Height, Width, Length)
• RoleProponent & Lead Developer