Return Portuguese meal type based on an image.
See https://www.kaggle.com/code/dima806/portuguese-meals-image-detection-vit for more details.
Classification report:
precision recall f1-score support
donuts 1.0000 0.9861 0.9930 216
hamburguer 1.0000 0.9954 0.9977 216
feijoada 0.9954 0.9908 0.9931 217
batatas_fritas 1.0000 1.0000 1.0000 216
esparguete_bolonhesa 1.0000 1.0000 1.0000 216
caldo_verde 0.9954 1.0000 0.9977 217
pasteis_bacalhau 0.9954 1.0000 0.9977 217
cozido_portuguesa 1.0000 1.0000 1.0000 216
jardineira 1.0000 1.0000 1.0000 217
arroz_cabidela 1.0000 1.0000 1.0000 216
nata 1.0000 1.0000 1.0000 216
croissant 1.0000 1.0000 1.0000 216
cachorro 0.9954 0.9954 0.9954 217
tripas_moda_porto 0.9909 1.0000 0.9954 217
aletria 0.9954 1.0000 0.9977 216
pizza 0.9954 0.9954 0.9954 217
bacalhau_natas 1.0000 1.0000 1.0000 216
ovo 0.9954 1.0000 0.9977 217
waffles 1.0000 1.0000 1.0000 216
francesinha 1.0000 1.0000 1.0000 217
bolo_chocolate 1.0000 0.9954 0.9977 216
gelado 0.9954 0.9954 0.9954 217
bacalhau_bras 1.0000 1.0000 1.0000 216
accuracy 0.9980 4978
macro avg 0.9980 0.9980 0.9980 4978
weighted avg 0.9980 0.9980 0.9980 4978
- Downloads last month
- 12
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Model tree for dima806/portuguese_meals_image_detection
Base model
google/vit-base-patch16-224-in21k