Trained a 15-CNN ensemble achieving 96.64% accuracy across 6 sugarcane disease classes, enabling early detection from leaf photos to reduce crop loss.
Deployed a full-stack web app with a FastAPI inference server for real-time classification.
AsukAI
Sugarcane disease classifier with a 15-CNN ensemble at 96.64% accuracy.
PythonPyTorchFastAPICNNHuggingFace