Automatic food detection in egocentric images using artificial intelligence technology

• Published in: Public health nutrition Vol. 22; no. 7; pp. 1168 - 1179
• Main Authors: Jia, Wenyan, Li, Yuecheng, Qu, Ruowei, Baranowski, Thomas, Burke, Lora E, Zhang, Hong, Bai, Yicheng, Mancino, Juliet M, Xu, Guizhi, Mao, Zhi-Hong, Sun, Mingui
• Format: Journal Article
• Language: English
• Published: Cambridge, UK Cambridge University Press 01.05.2019
• Subjects: Activities of Daily Living; Algorithms; Artificial Intelligence; Artificial neural networks; Cameras; Cooking; data collection; Data processing; Datasets; Deep learning; Diet Records; Dietetics - instrumentation; exercise; Food; Food quality; foods; Gardening; HOT TOPIC: ICT Assisted Dietary Data Collection and Analysis; Humans; Image acquisition; Image detection; Image Processing, Computer-Assisted; Image quality; information processing; lifestyle; Neural networks; nutrition assessment; Photography - instrumentation; Privacy; Recording; Research paper; restaurants; Sensitivity; Sensors; Wearable technology; New York; United States > US; Switzerland; Deep learning; Wearable device; Technology-assisted dietary intakeestimation; Food detection; Artificial intelligence; Egocentric image
• ISSN: 1368-9800; 1475-2727; 1475-2727
• DOI: 10.1017/S1368980018000538

To develop an artificial intelligence (AI)-based algorithm which can automatically detect food items from images acquired by an egocentric wearable camera for dietary assessment. To study human diet and lifestyle, large sets of egocentric images were acquired using a wearable device, called eButton, from free-living individuals. Three thousand nine hundred images containing real-world activities, which formed eButton data set 1, were manually selected from thirty subjects. eButton data set 2 contained 29 515 images acquired from a research participant in a week-long unrestricted recording. They included both food- and non-food-related real-life activities, such as dining at both home and restaurants, cooking, shopping, gardening, housekeeping chores, taking classes, gym exercise, etc. All images in these data sets were classified as food/non-food images based on their tags generated by a convolutional neural network. A cross data-set test was conducted on eButton data set 1. The overall accuracy of food detection was 91·5 and 86·4 %, respectively, when one-half of data set 1 was used for training and the other half for testing. For eButton data set 2, 74·0 % sensitivity and 87·0 % specificity were obtained if both 'food' and 'drink' were considered as food images. Alternatively, if only 'food' items were considered, the sensitivity and specificity reached 85·0 and 85·8 %, respectively. The AI technology can automatically detect foods from low-quality, wearable camera-acquired real-world egocentric images with reasonable accuracy, reducing both the burden of data processing and privacy concerns.