Being a Supercook: Joint Food Attributes and Multimodal Content Modeling for Recipe Retrieval and Exploration

doi:10.1109/TMM.2016.2639382

CORC > 自动化研究所 > 中国科学院自动化研究所 > 模式识别国家重点实验室 > 多媒体计算与图形学团队

	Being a Supercook: Joint Food Attributes and Multimodal Content Modeling for Recipe Retrieval and Exploration
	Min, Weiqing 1; Jiang, Shuqiang 1; Sang, Jitao2 ; Wang, Huayang 1; Liu, Xinda 3; Herranz, Luis 1
刊名	IEEE TRANSACTIONS ON MULTIMEDIA
	2017-05-01
卷号	19 期号:5 页码:1100-1113
关键词	Cuisine Classification Recipe Image Retrieval Ingredient Inference Multitask Deep Belief Network
DOI	10.1109/TMM.2016.2639382
文献子类	Article
英文摘要	This paper considers the problem of recipe-oriented image-ingredient correlation learning with multi-attributes for recipe retrieval and exploration. Existing methods mainly focus on food visual information for recognition while we model visual information, textual content (e.g., ingredients), and attributes (e.g., cuisine and course) together to solve extended recipe-oriented problems, such as multimodal cuisine classification and attributeenhanced food image retrieval. As a solution, we propose a multimodal multitask deep belief network (M3TDBN) to learn joint image-ingredient representation regularized by different attributes. By grouping ingredients into visible ingredients (which are visible in the food image, e.g., "chicken" and "mushroom") and nonvisible ingredients (e. g., "salt" and "oil"), M3TDBN is capable of learning both midlevel visual representation between images and visible ingredients and nonvisual representation. Furthermore, in order to utilize different attributes to improve the intermodality correlation, M3TDBN incorporates multitask learning to make different attributes collaborate each other. Based on the proposed M3TDBN, we exploit the derived deep features and the discovered correlations for three extended novel applications: 1) multimodal cuisine classification; 2) attribute-augmented cross-modal recipe image retrieval; and 3) ingredient and attribute inference fromfood images. The proposed approach is evaluated on the constructed Yummly dataset and the evaluation results have validated the effectiveness of the proposed approach.
WOS关键词	BOLTZMANN MACHINES
WOS研究方向	Computer Science ; Telecommunications
语种	英语
WOS记录号	WOS:000404056000017
资助机构	National Natural Science Foundation of China(61322212 ; National High Technology Research and Development 863 Program of China(2014AA015202) ; Beijing Municipal Commission of Science and Technology(D161100001816001) ; Lenovo Outstanding Young Scientists Program ; National Program for Special Support of Eminent Professionals ; China Post-doctoral Science Foundation(2016M590135) ; National Program for Support of Top-Notch Young Professionals ; 61532018 ; 61550110505 ; 61602437 ; 61373122)
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/15237]
专题	自动化研究所_模式识别国家重点实验室_多媒体计算与图形学团队
作者单位	1.Chinese Acad Sci, Inst Comp Technol, Key Lab Intelligent Informat Proc, Beijing 100190, Peoples R China 2.Chinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China 3.Beihang Univ, State Key Lab Virtual Real Technol & Syst, Beijing 100191, Peoples R China
推荐引用方式 GB/T 7714	Min, Weiqing,Jiang, Shuqiang,Sang, Jitao,et al. Being a Supercook: Joint Food Attributes and Multimodal Content Modeling for Recipe Retrieval and Exploration[J]. IEEE TRANSACTIONS ON MULTIMEDIA,2017,19(5):1100-1113.
APA	Min, Weiqing,Jiang, Shuqiang,Sang, Jitao,Wang, Huayang,Liu, Xinda,&Herranz, Luis.(2017).Being a Supercook: Joint Food Attributes and Multimodal Content Modeling for Recipe Retrieval and Exploration.IEEE TRANSACTIONS ON MULTIMEDIA,19(5),1100-1113.
MLA	Min, Weiqing,et al."Being a Supercook: Joint Food Attributes and Multimodal Content Modeling for Recipe Retrieval and Exploration".IEEE TRANSACTIONS ON MULTIMEDIA 19.5(2017):1100-1113.