Achieved 75% accuracy for a validation dataset for classifying 80 types of common Indian food.
See my Kaggle notebook for more details.
Classification report:
precision recall f1-score support
adhirasam 0.9412 0.8000 0.8649 20
aloo_gobi 0.7857 0.5500 0.6471 20
aloo_matar 0.8500 0.8500 0.8500 20
aloo_methi 0.7407 1.0000 0.8511 20
aloo_shimla_mirch 0.7619 0.8000 0.7805 20
aloo_tikki 1.0000 0.7500 0.8571 20
anarsa 1.0000 0.7000 0.8235 20
ariselu 0.7692 1.0000 0.8696 20
bandar_laddu 0.8333 0.7500 0.7895 20
basundi 0.2254 0.8000 0.3516 20
bhatura 0.7600 0.9500 0.8444 20
bhindi_masala 0.8636 0.9500 0.9048 20
biryani 0.8571 0.9000 0.8780 20
boondi 0.9474 0.9000 0.9231 20
butter_chicken 0.4419 0.9500 0.6032 20
chak_hao_kheer 0.9474 0.9000 0.9231 20
cham_cham 1.0000 0.4000 0.5714 20
chana_masala 0.7692 1.0000 0.8696 20
chapati 0.7407 1.0000 0.8511 20
chhena_kheeri 0.0000 0.0000 0.0000 20
chicken_razala 0.8000 1.0000 0.8889 20
chicken_tikka 0.9091 0.5000 0.6452 20
chicken_tikka_masala 0.7273 0.4000 0.5161 20
chikki 0.7308 0.9500 0.8261 20
daal_baati_churma 0.6957 0.8000 0.7442 20
daal_puri 1.0000 0.3000 0.4615 20
dal_makhani 0.8182 0.9000 0.8571 20
dal_tadka 0.6552 0.9500 0.7755 20
dharwad_pedha 1.0000 0.8000 0.8889 20
doodhpak 0.6667 0.1000 0.1739 20
double_ka_meetha 0.7917 0.9500 0.8636 20
dum_aloo 0.8462 0.5500 0.6667 20
gajar_ka_halwa 0.8000 1.0000 0.8889 20
gavvalu 0.8095 0.8500 0.8293 20
ghevar 1.0000 0.8000 0.8889 20
gulab_jamun 0.5429 0.9500 0.6909 20
imarti 0.8333 1.0000 0.9091 20
jalebi 0.9474 0.9000 0.9231 20
kachori 0.6364 0.7000 0.6667 20
kadai_paneer 0.6923 0.9000 0.7826 20
kadhi_pakoda 0.8500 0.8500 0.8500 20
kajjikaya 0.9412 0.8000 0.8649 20
kakinada_khaja 0.8824 0.7500 0.8108 20
kalakand 0.7692 0.5000 0.6061 20
karela_bharta 1.0000 0.2000 0.3333 20
kofta 0.9333 0.7000 0.8000 20
kuzhi_paniyaram 0.6667 0.9000 0.7660 20
lassi 0.8000 1.0000 0.8889 20
ledikeni 0.5714 0.2000 0.2963 20
litti_chokha 1.0000 0.8000 0.8889 20
lyangcha 0.8947 0.8500 0.8718 20
maach_jhol 0.9375 0.7500 0.8333 20
makki_di_roti_sarson_da_saag 1.0000 0.8500 0.9189 20
malapua 1.0000 0.7000 0.8235 20
misi_roti 0.8571 0.9000 0.8780 20
misti_doi 0.6364 0.7000 0.6667 20
modak 0.7826 0.9000 0.8372 20
mysore_pak 0.7917 0.9500 0.8636 20
naan 0.9091 1.0000 0.9524 20
navrattan_korma 0.9286 0.6500 0.7647 20
palak_paneer 0.7917 0.9500 0.8636 20
paneer_butter_masala 0.6667 0.7000 0.6829 20
phirni 0.5500 0.5500 0.5500 20
pithe 1.0000 0.2500 0.4000 20
poha 0.6786 0.9500 0.7917 20
poornalu 0.9000 0.9000 0.9000 20
pootharekulu 0.8636 0.9500 0.9048 20
qubani_ka_meetha 1.0000 0.6500 0.7879 20
rabri 0.0000 0.0000 0.0000 20
ras_malai 0.7083 0.8500 0.7727 20
rasgulla 0.5263 1.0000 0.6897 20
sandesh 0.6000 0.1500 0.2400 20
shankarpali 0.8333 1.0000 0.9091 20
sheer_korma 0.4643 0.6500 0.5417 20
sheera 0.8667 0.6500 0.7429 20
shrikhand 0.8000 0.6000 0.6857 20
sohan_halwa 1.0000 0.5000 0.6667 20
sohan_papdi 0.5556 1.0000 0.7143 20
sutar_feni 0.8571 0.9000 0.8780 20
unni_appam 0.5556 0.7500 0.6383 20
accuracy 0.7519 1600
macro avg 0.7813 0.7519 0.7352 1600
weighted avg 0.7813 0.7519 0.7352 1600
- Downloads last month
- 49
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/indian_food_image_detection
Base model
google/vit-base-patch16-224-in21k