metadata
tags:
- image-classification
- pytorch
- huggingpics
metrics:
- accuracy
model-index:
- name: dog-races-v2
results:
- task:
name: Image Classification
type: image-classification
metrics:
- name: Accuracy
type: accuracy
value: 0.8299999833106995
dog-races-v2
Autogenerated Model created thannks to HuggingPics🤗🖼️. You can create your own image classifier for anything by running the demo on Google Colab.
This Model is an improvement to my last model, where the Chow Chow data included images of American pickles with the same name (contaminated data). Current labels are: 1) Border Collie, 2) Chow Chow, 3) German Shepherd, 4) Golden Retriever, 5) Pug, 6) Rottweiler, 7) Shiba Inu, 8) Siberian Husky and 9) Tibetan Mastiff.
There is still room for improvement.
Model Accuracy: 82.99%
When tested with Stanford Dogs Dataset, these were the results:
- Golden Retriever: 90% (117/130 images labeled correctly)
- Chow Chow: 97.45% (191/196 images labeled correctly)
- Tibetan Mastiff: 12.5% (19/152 images labeled correctly). Probably some issue with the data (most were labeled as Chow Chow).
Example Images
Border Collie
Chow Chow dog
German Shepherd
Golden Retriever
Pug
Rottweiler
Shiba Inu
Siberian Husky
Tibetan Mastiff
