lakshitha1629/Canine-Skin-Image-Classifier

Canine Skin Image Classifier

This repository mirrors the Keras checkpoint and labels from the Pet Disease Detection GitHub project. The project's notebook uses it to classify dog images as CanineImpetigo or CanineLupus.

Files

• keras_model.h5: original Keras model from Keras AI Part/keras_model.h5 (2.45 MB)
• labels.txt: class index 0 = CanineImpetigo; class index 1 = CanineLupus

Input and output

The source notebook resizes an RGB image to 224 × 224, converts it to a float32 array, and normalizes pixel values with (pixel / 127.0) - 1. It feeds an array of shape (1, 224, 224, 3) into the model and interprets output index 1 above 0.5 as CanineLupus. This description follows the source code; independent inference testing has not been performed here.

Limitations

The source repository does not document the training images' provenance, dataset size, evaluation metrics, or performance on independent images. This is a project demonstration and must not be used to diagnose an animal or choose treatment. Consult a veterinarian for health decisions. Keras/HDF5 model loading may depend on the library versions used when the file was saved.

Original files: checkpoint · labels.