Food-101

Benchmarks

Task Name	Dataset Name	SOTA Result
Multimodal Multiclass Classification	FOOD-101 (test)	Accuracy94.6	45
Image Classification	Food-101 LT	Avg Class Recall49	12
Fine-grained Visual Categorization	Food-101 (test)	Accuracy98.6	11
Label Noise Detection	Food-101N (val)	Average Error Rate6.99	11
Image Classification	Food-101N r ≈ 20% (test)	Accuracy86.4	10
Label Noise Detection	Food-101N verified (val)	Error Rate6.99	10
Clustering	Food-101	ARI0.478	9
Image Classification	Food-101 16K sequence length (test)	Train runtime (s)891	7
Image Classification	Food-101 all-to-all	Accuracy86.3	7
Out-of-Distribution Detection	Food-101 (test)	iNaturalist FPR951.31	7
Multimodal Classification	FOOD-101 Hard (test)	Accuracy93.4	7
Multimodal Classification	UPMC Food-101 (val)	Accuracy80.39	5
Image Classification	Food-101N 25k (test)	Top-1 Accuracy85.78	5
Classification	FOOD-101	Accuracy81.68	4
Image Classification	Food-101 8-shot	Accuracy91.4	3
Image Classification	Food-101 Red Meat (test)	Accuracy71.2	3
Image Classification	Food-101 (low-confidence set)	Top-1 Acc49	2
recipe2im	Food-101 (test)	Median Rank (medR)2.6	2
im2recipe	Food-101 (test)	Median Rank10.15	2
Image Recognition	Food-101	Top-1 Acc97.4	1
Visual Recognition	Food-101	Top-1 Acc99.3	1

Showing 21 of 21 rows