🐾 StreetAnimalTriageAI: Multi-Layered Cascading Vision Network for Emergency Veterinary Assessment
StreetAnimalTriageAI is a high-precision, 3-stage cascading deep learning pipeline designed to automate the detection, geometric analysis, and pathological triage of street animals (dogs, cats, cows, sheep) in urban environments.
Built completely from scratch using custom-trained YOLOv8 paradigms, the platform isolates target assets, outlines soft-tissue injuries using instance segmentation, measures the exact structural body defect ratio, and generates species-safe first-aid protocols alongside emergency dispatch triggers.
🏗️ System Architecture & Cascading Logic
Rather than relying on a single monolithic vision model, this system splits the complex diagnostic problem into three specialized, highly efficient sequential layers. Each layer optimizes resources by processing cropped, contextual target regions passed down by its predecessor.