A Benchmark Dataset to Study the Representation of Food Images

Farinella, Giovanni Maria; Allegra, Dario; Stanco, Filippo

doi:10.1007/978-3-319-16199-0_41

Cited by 63 publications

(46 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this paper we propose a novel method for matching food images, that is conceptually more similar to [1] rather than to the other papers discussed in Section II. Therefore, no classifiers have been built for the known food categories, and the method is open to unknown types of food without any modification or retraining .…”

Section: A Methodologigal Principlesmentioning

confidence: 99%

See 1 more Smart Citation

Machine Vision in Food Recognition: Attempts to Enhance CBVIR Tools

Sluzek

2016

Annals of Computer Science and Information Systems

View full text Add to dashboard Cite

Abstract-Visual identification of complex images (e.g. images of food) remains a challenging problem. In particular, contentbased visual information retrieval (CBVIR) methods, which seem a natural choice for such tasks, are often constrained by specific characteristics of the images of interest and (possibly) other practical requirements. In this paper, a novel CBVIR approach to automatic food identification is proposed, taking into account characteristics of solutions currently existing in this area. Based on limitations of those solutions, we present a scheme in which a co-occurrence of MSER features extracted from three color channels is employed to build a bag-of-words histogram. Subsequently, food images are matched by detecting similarities between those histograms. Preliminary tests on a recently published benchmark dataset UNICT-FD889 reveal certain advantages of the scheme and highlight its limitations. In particular, a need of a novel methodology for segmentation of food images has been identified.

show abstract

Section: A Methodologigal Principlesmentioning

confidence: 99%

“…Currently, it seems the most comprehensive (in terms of the number of food items) study was presented in [1]. UNICT-FD889 dataset of almost 900 diversified dishes (see examples in Fig.…”

Section: Vision-based Techniques For Food Recognitionmentioning

confidence: 99%

Machine Vision in Food Recognition: Attempts to Enhance CBVIR Tools

Sluzek

2016

Annals of Computer Science and Information Systems

View full text Add to dashboard Cite

show abstract

“…There are several publicly available food datasets that only provide labels of the food types that appear in the image, they do not provide any information relative to portion size (i.e., segmented regions and food volume) that we also require. These publicly available datasets include the 100,000+ images in the Food-101 dataset [12], approximately 1 million food images in the ImageNet dataset [32], the 4545 images and 606 stereo pairs in the Pittsburgh Fast Food image dataset [13,47], the 5000 images in the Foodcam Dataset [25, 26, 34], the 3000 images in the FooDD dataset [38], and the 3583 images in the UNICT-FD889 dataset [17]. …”

Section: Introductionmentioning

confidence: 99%

A Mobile Food Record For Integrated Dietary Assessment

Ahmad

Bosch

Khanna

et al. 2016

Proceedings of the 2nd International Workshop on Multimedia Assisted Dietary Management

View full text Add to dashboard Cite

This paper presents an integrated dietary assessment system based on food image analysis that uses mobile devices or smartphones. We describe two components of our integrated system: a mobile application and an image-based food nutrient database that is connected to the mobile application. An easy-to-use mobile application user interface is described that was designed based on user preferences as well as the requirements of the image analysis methods. The user interface is validated by user feedback collected from several studies. Food nutrient and image databases are also described which facilitates image-based dietary assessment and enable dietitians and other healthcare professionals to monitor patients dietary intake in real-time. The system has been tested and validated in several user studies involving more than 500 users who took more than 60,000 food images under controlled and community-dwelling conditions.

show abstract

“…They can recognize places and faces and display information and social media profiles of people the smart glasses user is looking at. Smart-glasses are already applied in medicine, and are envisioned to have a wide range of applications in this field Moshtaghi et al 2015;Hetterich et al 2014;Mentler et al 2015;Klein et al 2015; S. Maas et al 2015;Rankin et al 2015), education (Ikonen and Knutas ;Labus et al 2015;Freina and Ott 2015), tourism (Harasymowicz 2015), social science research (Paterson and Glass 2015), navigation (Ostendorp et al 2015;Higuchi et al 2015), crowd steering (Borean et al 2015), activity recognition (Zhan 2014;Betancourt et al 2015) including diet recognition and food behavior control (Gemming et al 2013;Farinella et al 2014), mood (engagement) measurement (Kunze et al 2015), forensics (Karabiyik 2015), promoting cultural sustainability (Irving and Hoffman 2014), promoting teamwork and safety (Moshtaghi et al 2015) and they can also be used for military purposes. Prototypes and applications exist in several of these fields.…”

Section: Introductionmentioning

confidence: 99%