Food101

Benchmarks

Task Name	Dataset Name	SOTA Result
Image Classification	Food101 (test)	Accuracy96.38	91
Image Classification	Food101	Robust Accuracy96.11	49
Class-Incremental Learning	Food101-LT rho=100 (test)	Accuracy36.84	48
Image Classification	Food101 Base-to-new generalization	Accuracy (Base Classes)93.75	47
Category Discovery	Food101	Accuracy (All)61.2	45
Fine-grained Classification	Food101	Top-1 Acc93	42
Multimodal Classification	FOOD101 UPMC (train test)	Accuracy94.2	36
Image Classification	Food101	Top-1 Accuracy96.38	33
Image-to-image retrieval	Food101	mAP71	30
Image Classification	Food101 novel classes	ECE0.0074	29
Error Detection	Food101	AuROC99.92	29
Fine-grained classification	Food101 (base classes)	Accuracy91.71	27
OOD Detection	Food101	AUROC92.63	19
Image Classification	Food101	AULC77.9	18
Image Classification	Food101 H (harmonic mean)	Accuracy91.38	16
Image Classification	Food101 Base	Top-1 Accuracy91.63	16
Fine-grained classification	Food101 novel classes	MCE0.25	15
Image Classification	Food101 5-way 5-shot (test)	Accuracy (%)79.59	13
Image Classification	Food101 5-way 2-shot	Accuracy64.18	13
Image Classification	Food101 New classes	Accuracy92.05	12
Fine-grained category recognition and content assessment	Food101	Score75	12
Image Classification	Food101N	Accuracy85.61	12
Class-Incremental Learning	Food101-LT 10-task setting	Average Mean Accuracy59.38	10
Class-Incremental Learning	Food101-LT 5-task setting	Average Mean Accuracy (AMean)65.19	10
Class-Incremental Learning	Food101 10-task setting (test)	Average Mean Accuracy74.57	10

Showing 25 of 61 rows