mopineyro / resnet_breeds_finetuned

ResNet Fine-Tuned on 37 dog & cat breeds

  • Public
  • 133 runs

Input

Output

Run time and cost

This model runs on CPU hardware. Predictions typically complete within 1 seconds. The predict time for this model varies significantly based on the inputs.

Readme

Model Details

The model is based on the ResNet18 architecture, adapted to classify 37 different breeds of dogs and cats. The final layer of the model has been modified to output 37 classes, corresponding to these breeds.

The model expects input images in the form of URLs, retrieves them, preprocesses them to fit the model requirements, and then forwards them through the network to obtain a prediction.

Image Preprocessing

The image preprocessing steps include resizing the image to a fixed dimension (224x224 pixels), converting it to a tensor, and normalizing it with specific mean and standard deviation values used for ResNet models.

Error Handling

The model includes error handling for scenarios where the image URL might be incorrect or the image cannot be retrieved. It logs appropriate error messages to assist in debugging.