Table 1.

Overview of notable food identification systems, classifier algorithms, selected features, number of classes, name of food dataset (if specified or noted as their own dataset if absent), and accuracy results^a.

Study, year	Projects or team	Classifier	Feature	Class (dataset)	Accuracy results percentages
Shroff et al [18], 2008	DiaWear	Neural network	Color, size, shape, and texture	4	~75
Chen et al [19], 2009	PFIDb	SVMc	Color BoSIFTd	61 (PFID)	~11 ~24
Taichi and Keiji [21], 2009	UECe	MKLf	Color, texture, and SIFTg	50	61.34
Hoashi et al [52], 2010	UEC	MKL	BoFh, Gabori, color, HOGj, and texture	85	62.53
Yang et al [20], 2010	PFID	SVM	Pairwise local features	61 (PFID)	78.00
Zhu et al [31], 2010	TADAk	SVM with Gaussian radial basis kernel	Color and texture	19	97.20
Kong and Tan [53], 2011	DietCam	Multiclass SVM	Nearest neighbor Gaussian region detector, and SIFT	61 (PFID)	84.00
Bosch et al [22], 2011	TADA	SVM	Color, entropy, Gabor, Tamural, SIFT, Haar waveletm, steerablen, and DAISYo	39	86.10
Matsuda et al [36], 2012	UEC	MKL-SVM	HOG, SIFT, Gabor, color, and texture	100 (UEC-Food100)	55.80
Anthimopoulos et al [23], 2014	GoCARB	SVM	HSVp-SIFT, optimized BoF, and color moment invariant	11	78.00
He et al [54], 2014	TADA	k-nearest neighbors	DCDq, SIFT, MDSIFTr, and SCDs	42	65.4
Pouladzadeh et al [32], 2014	—t	SVM	Color, texture, size, and shape	15	90.41
Kawano and Yanai [35], 2014	UEC	Pretrained CNNu	—	100 (UEC-Food100)	72.3
Yanai and Kawano [39], 2015	UEC	Deep CNN	—	100 (UEC-Food-100)	78.77
Christodoulidis et al [40], 2015	GoCARB	Patch-wise CNN	—	7	84.90
Myers et al [27], 2015	Google	GoogLeNet	—	101	79.00
Liu et al [41], 2016	—	DeepFood	—	Food-101 UEC-256	77.40 54.70
Singla et al [42], 2016	—	GoogLeNet	—	11	83.60
Hassannejad et al [43], 2016	—	InceptionV3v	—	101 (Food-101) 100 (UEC-Food100) 256 (UEC-Food256)	88.28 81.45 76.17
Ciocca et al [44], 2017	—	VGGw	—	73 (UNIMINB2016)	78.30
Mezgec and Koroušić Seljak [45], 2017	—	NutriNet (Modified AlexNetx)	—	73 (UNIMINB2016)	86.72
Pandey et al [55], 2017	—	Ensemble net	—	101 (Food-101)	72.10
Martinel et al [56], 2018	—	WISeRy	—	101 (Food-101) 100 (UEC-Food100) 256 (UEC-Food256)	88.72 79.76 86.71
Jiang et al [57], 2020	—	MSMVFAz	—	101 (Food-101) 172 (VireoFood-172) 208 (ChineseFoodNet)	~90.47 90.61 81.94
Lu et al [58], 2020	GoCARB	Modified InceptionV3	—	298 Generic food Subgroups Fine-grained (MADiMAaa)	65.80 61.50 57.10
Wu et al [59], 2021	—	Modified AlexNet	—	22 styles of Bento sets	96.30

^aNote that convolutional neural network–based classifiers do not require the number of features to be shown as they extract features autonomously.

^bPFID: Pittsburgh Fast-Food Image Dataset.

^cSVM: support vector machine.

^dBoSIFT: bag-of-scale-invariant feature transform.

^eUEC: University of Electro-Communications.

^fMKL: multiple kernel learning. This is a machine-learning technique that combines multiple kernels or similarity functions, to improve the performance and flexibility of kernel-based models such as support vector machines.

^gSIFT: scale-invariant feature transform.

^hBoF: bag-of-features.

ⁱGabor is a texture feature extraction invented by Dennis Gabor.

^jHOG: histogram of orientated gradients—a feature descriptor based on color.

^kTADA: Technology Assisted Dietary Assessment.

^lTamura is a 6-texture feature extraction invented by Hideyuki Tamura.

^mHaar wavelet is a mathematical analysis for wavelet sequence named after Alfréd Haar.

ⁿSteerable filter is an image filter introduced by Freeman and Adelson.

^oDAISY is a local image descriptor introduced by E Tola et al [60], but they did not describe a true acronym of DAISY.

^pHSV is the name of a red-green-blue color model based on hue, saturation, and value.

^qDCD: dominant color descriptor.

^rMDSIFT: multiscale dense scale-invariant feature transform.

^sSCD: scalable color descriptor.

^tNot available.

^uCNN: convolutional neural network.

^vInception is an object detection model that won the ImageNet Challenge in 2014, recognized for its use of a novel architecture that efficiently leverages computing resources inside the network.

^wVGG: visual geometry group—an object detection model named after a research group from the University of Oxford.

^xAlexNet is an object detection model that won the ImageNet Large-Scale Visual Recognition Challenge (also known as the ImageNet challenge) in 2012; it is named after its inventors, Alex Krizhevsky.

^yWISeR: wide-slice residual.

^zMSMVFA: multi-scale multi-view feature aggregation.

^aaMADiMA: Multimedia Assisted Dietary Management.