Food Images Research Articles

Socio-cultural factors contributing to eating disorder development amongst women of South Asian ethnicity: a meta-synthesis

Purpose The purpose of this study is to synthesize published empirical qualitative evidence which explores the socio-cultural influences on the development of eating disorders from the perspective of South Asian women. Design/methodology/approach A systematic search of the databases CINAHL, MEDLINE, Scopus and APA PsycINFO was conducted to identify relevant journal articles published up to August 2023. The quality of the included studies was appraised using the Critical Appraisal Skills Programme Qualitative Tool. Data was synthesized and analysed using thematic synthesis. Findings Four studies from the UK, USA and Canada were included in the meta-synthesis. Three main themes were found: gratifying others, negative childhood experiences and shame and stigma. Practical implications South Asian women report many socio-cultural factors that can influence eating disorder development. Findings of this meta-synthesis highlight the importance of cultural awareness, and health-care professionals should be aware of the nuanced socio-cultural factors that can influence a South Asian woman’s relationship with food and body image. Originality/value South Asian communities are understudied in the context of mental health, including eating disorders. To the best of the authors’ knowledge, this is the first study to synthesize the available qualitative evidence on socio-cultural influences on eating disorder development among South Asian women. These findings have revealed insights that can be harnessed by health-care professionals to better understand factors that contribute to eating disorder development and support cultural awareness.

Achieving an Improved NIR Performance of Ca4-xSc2xZr1-xGe3O12:Cr3+ via [Sc3+-Sc3+]→[Ca2+-Zr4+

High-performance near-infrared (NIR) light sources are highly sought after in advanced spectroscopy techniques, driving the development of NIR phosphor-converted light-emitting diodes (pc-LEDs). Escalating the luminescence intensity and thermal stability of NIR-emitting phosphors, which is a core component of NIR pc-LEDs, is of paramount importance. Herein, chemical unit cosubstitution and cosolvent addition tactics were implemented to simultaneously boost the NIR luminescence performance of the synthesized phosphors. The replacement of [Sc3+-Sc3+] for [Ca2+-Zr4+] in Ca4ZrGe3O12:Cr3+ likely reduces the antisite defects and offers a more rigid crystal structure. As a result, the emitting intensity is reinforced significantly, along with a remarkable improvement in thermal stability, acquiring 65% of the initial luminescence at 423 K compared with 42% for the primal sample. Moreover, the introduction of H3BO3 further enhances NIR luminescence while maintaining a favorable thermal resistance. The encapsulated NIR pc-LED carries an impressive output power of 71.8 mW at 300 mA and a conversion efficiency up to 15.6% at 10 mA. The practical presentations in food checking, imaging, and detection manifest that Ca3.5ScZr0.5Ge3O12:Cr3+,H3BO3 is a promising material for spectroscopy-based technologies.

Seeing is believing: How front-of-pack food imagery shapes parents' perceptions of child-oriented snack foods.

Child-oriented snack foods often display front-of-pack images suggesting they are natural or wholesome, yet many of these products are ultra-processed and nutrient poor. This study investigated parent's perceptions in response to common forms of front-of-pack food imagery on child-orientated snacks. An online experiment comprising four between-subjects food image conditions (no image; food photo; food cartoon; cartoon of children gardening) and two within-subjects snack type conditions (fruit; vegetable) was conducted with N=800 Australian adult parents/guardians of children aged 4-10 years. Participants viewed packaged snacks then rated them on visual appeal, perceived healthfulness, suitability for their child, and purchase intentions. For each outcome, 4 (between-subjects) x 2 (within-subjects) mixed ANOVAs were performed, with significant main effects explored using post-hoc pairwise t-tests with Bonferroni adjustment. The extent to which perception metrics predicted purchase intention within each condition was examined using multiple regression analyses. Participants perceived snacks displaying food photographs as healthier compared to all other conditions. When children's snacks displayed food photos, participants perceived the snack as more suitable for their child and showed stronger intentions to purchase the product than when cartoon children were displayed. Participants preferred fruit over vegetable snacks across all metrics, particularly when the picture was photorealistic. Visual appeal was the only perception metric that consistently predicted purchase intention across conditions. Findings provide clear evidence that parent's perceptions and purchase intentions towards children's snack foods are enhanced by pictures displayed on the packaging. Policy to regulate food packaging should require that foods depicted on the packaging accurately reflect product ingredients, particularly where whole foods are represented, and the actual ingredients are highly processed and no longer provide the nutrition of whole foods.

FoodCSWin: A High-Accuracy Food Image Recognition Model for Dietary Assessment

FoodCSWin: A High-Accuracy Food Image Recognition Model for Dietary Assessment

Tourist Loyalty in Bali Tourism: Can Tourist Satisfaction Mediate the Effect of Quality of Food and Destination Image?

Tourist Loyalty in Bali Tourism: Can Tourist Satisfaction Mediate the Effect of Quality of Food and Destination Image?

Artificial Intelligence Applications to Measure Food and Nutrient Intakes: Scoping Review.

Accurate measurement of food and nutrient intake is crucial for nutrition research, dietary surveillance, and disease management, but traditional methods such as 24-hour dietary recalls, food diaries, and food frequency questionnaires are often prone to recall error and social desirability bias, limiting their reliability. With the advancement of artificial intelligence (AI), there is potential to overcome these limitations through automated, objective, and scalable dietary assessment techniques. However, the effectiveness and challenges of AI applications in this domain remain inadequately explored. This study aimed to conduct a scoping review to synthesize existing literature on the efficacy, accuracy, and challenges of using AI tools in assessing food and nutrient intakes, offering insights into their current advantages and areas of improvement. This review followed the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines. A comprehensive literature search was conducted in 4 databases-PubMed, Web of Science, Cochrane Library, and EBSCO-covering publications from the databases' inception to June 30, 2023. Studies were included if they used modern AI approaches to assess food and nutrient intakes in human subjects. The 25 included studies, published between 2010 and 2023, involved sample sizes ranging from 10 to 38,415 participants. These studies used a variety of input data types, including food images (n=10), sound and jaw motion data from wearable devices (n=9), and text data (n=4), with 2 studies combining multiple input types. AI models applied included deep learning (eg, convolutional neural networks), machine learning (eg, support vector machines), and hybrid approaches. Applications were categorized into dietary intake assessment, food detection, nutrient estimation, and food intake prediction. Food detection accuracies ranged from 74% to 99.85%, and nutrient estimation errors varied between 10% and 15%. For instance, the RGB-D (Red, Green, Blue-Depth) fusion network achieved a mean absolute error of 15% in calorie estimation, and a sound-based classification model reached up to 94% accuracy in detecting food intake based on jaw motion and chewing patterns. In addition, AI-based systems provided real-time monitoring capabilities, improving the precision of dietary assessments and demonstrating the potential to reduce recall bias typically associated with traditional self-report methods. While AI demonstrated significant advantages in improving accuracy, reducing labor, and enabling real-time monitoring, challenges remain in adapting to diverse food types, ensuring algorithmic fairness, and addressing data privacy concerns. The findings suggest that AI has transformative potential for dietary assessment at both individual and population levels, supporting precision nutrition and chronic disease management. Future research should focus on enhancing the robustness of AI models across diverse dietary contexts and integrating biological sensors for a holistic dietary assessment approach.

Automated Nutrient Deficiency Detection and Recommendation Systems Using Deep Learning in Nutrition Science

Nutrient deficiencies affect millions globally, contributing to severe health issues and reduced quality of life. Traditional methods of diagnosing these deficiencies and recommending dietary adjustments are often time-intensive, prone to error, and lack personalization. The advent of deep learning has revolutionized nutrition science, offering automated, accurate, and scalable solutions. This paper delves into the development and application of automated nutrient deficiency detection and recommendation systems powered by deep learning. Key components of such systems include advanced data processing techniques that analyze multimodal datasets, such as biomarkers, dietary records, and food images. Convolutional Neural Networks (CNNs) excel in recognizing and quantifying nutrients from food images, while Recurrent Neural Networks (RNNs) handle time-series dietary data. Generative Adversarial Networks (GANs) and Natural Language Processing (NLP) facilitate data augmentation and textual analysis of dietary logs, respectively. These systems enable precise detection of deficiencies and generate tailored dietary plans based on individual needs, considering demographic and lifestyle factors. This article highlights case studies and practical implementations of deep learning models in real-world applications, such as AI-powered nutrition apps and biomarker-based deficiency prediction. It also addresses significant challenges, including data quality, algorithmic bias, and ethical concerns related to privacy and equity. Furthermore, the study explores future opportunities, such as integrating explainable AI, leveraging multi-modal data sources, and enhancing IoT-based tracking devices to improve recommendation systems. By bridging the gap between AI technology and nutrition science, these systems hold the potential to revolutionize global dietary health, offering scalable, personalized, and efficient solutions to combat nutrient deficiencies.

Multimodal Food Image Classification with Large Language Models

In this study, we leverage advancements in large language models (LLMs) for fine-grained food image classification. We achieve this by integrating textual features extracted from images using an LLM into a multimodal learning framework. Specifically, semantic textual descriptions generated by the LLM are encoded and combined with image features obtained from a transformer-based architecture to improve food image classification. Our approach employs a cross-attention mechanism to effectively fuse visual and textual modalities, enhancing the model’s ability to extract discriminative features beyond what can be achieved with visual features alone.

Advancements in Using AI for Dietary Assessment Based on Food Images: Scoping Review.

To accurately capture an individual's food intake, dietitians are often required to ask clients about their food frequencies and portions, and they have to rely on the client's memory, which can be burdensome. While taking food photos alongside food records can alleviate user burden and reduce errors in self-reporting, this method still requires trained staff to translate food photos into dietary intake data. Image-assisted dietary assessment (IADA) is an innovative approach that uses computer algorithms to mimic human performance in estimating dietary information from food images. This field has seen continuous improvement through advancements in computer science, particularly in artificial intelligence (AI). However, the technical nature of this field can make it challenging for those without a technical background to understand it completely. This review aims to fill the gap by providing a current overview of AI's integration into dietary assessment using food images. The content is organized chronologically and presented in an accessible manner for those unfamiliar with AI terminology. In addition, we discuss the systems' strengths and weaknesses and propose enhancements to improve IADA's accuracy and adoption in the nutrition community. This scoping review used PubMed and Google Scholar databases to identify relevant studies. The review focused on computational techniques used in IADA, specifically AI models, devices, and sensors, or digital methods for food recognition and food volume estimation published between 2008 and 2021. A total of 522 articles were initially identified. On the basis of a rigorous selection process, 84 (16.1%) articles were ultimately included in this review. The selected articles reveal that early systems, developed before 2015, relied on handcrafted machine learning algorithms to manage traditional sequential processes, such as segmentation, food identification, portion estimation, and nutrient calculations. Since 2015, these handcrafted algorithms have been largely replaced by deep learning algorithms for handling the same tasks. More recently, the traditional sequential process has been superseded by advanced algorithms, including multitask convolutional neural networks and generative adversarial networks. Most of the systems were validated for macronutrient and energy estimation, while only a few were capable of estimating micronutrients, such as sodium. Notably, significant advancements have been made in the field of IADA, with efforts focused on replicating humanlike performance. This review highlights the progress made by IADA, particularly in the areas of food identification and portion estimation. Advancements in AI techniques have shown great potential to improve the accuracy and efficiency of this field. However, it is crucial to involve dietitians and nutritionists in the development of these systems to ensure they meet the requirements and trust of professionals in the field.

Proto-DS: A Self-Supervised Learning-Based Nondestructive Testing Approach for Food Adulteration with Imbalanced Hyperspectral Data.

Conventional food fraud detection using hyperspectral imaging (HSI) relies on the discriminative power of machine learning. However, these approaches often assume a balanced class distribution in an ideal laboratory environment, which is impractical in real-world scenarios with diverse label distributions. This results in suboptimal performance when less frequent classes are overshadowed by the majority class during training. Thus, the critical research challenge emerges of how to develop an effective classifier on a small-scale imbalanced dataset without significant bias from the dominant class. In this paper, we propose a novel nondestructive detection approach, which we call the Dice Loss Improved Self-Supervised Learning-Based Prototypical Network (Proto-DS), designed to address this imbalanced learning challenge. The proposed amalgamation mitigates the label bias on the most frequent class, further improving robustness. We validate our proposed method on three collected hyperspectral food image datasets with varying degrees of data imbalance: Citri Reticulatae Pericarpium (Chenpi), Chinese herbs, and coffee beans. Comparisons with state-of-the-art imbalanced learning techniques, including the Synthetic Minority Oversampling Technique (SMOTE) and class-importance reweighting, reveal our method's superiority. Notably, our experiments demonstrate that Proto-DS consistently outperforms conventional approaches, achieving the best average balanced accuracy of 88.18% across various training sample sizes, whereas the Logistic Model Tree (LMT), Multi-Layer Perceptron (MLP), and Convolutional Neural Network (CNN) approaches attain only 59.42%, 60.38%, and 66.34%, respectively. Overall, self-supervised learning is key to improving imbalanced learning performance and outperforms related approaches, while both prototypical networks and the Dice loss can further enhance classification performance. Intriguingly, self-supervised learning can provide complementary information to existing imbalanced learning approaches. Combining these approaches may serve as a potential solution for building effective models with limited training data.

Maladaptive Restrictive Food Choice Behavior in Adolescents With Anorexia Nervosa.

Individuals with anorexia nervosa persistently restrict their food intake while often severely underweight. This maladaptive food choice behavior and related decision-making processes have mainly been investigated in adults. It is unknown whether the same decision-making processes drive food choices in adolescents, given their more favorable treatment outcomes. This study investigated maladaptive food choice behavior in adolescents with AN and examined whether they display the same decision-making processes as adults. Adolescents with AN (n = 42) and age-matched controls (n = 42) completed a computer-based food choice task, rating food images for healthiness and tastiness before choosing between two food items. Adolescents with AN chose high-fat foods less frequently than controls, with food choices more influenced by their perceived "healthiness" and less by their "tastiness" than they did among controls. Relative to controls, adolescents with AN also reported lower overall tastiness ratings and greater habit strength of restrictive eating, which was, however, not related to food choices. Adolescents with AN display the same maladaptive food choice behavior as adults, namely the persistent restriction of high-fat food choices. Their choices were more strongly influenced by the perceived healthiness of a food item, compared to the choices of controls.

Understanding childhood obesity in Pakistan: exploring the knowledge, attitudes, practices of mothers, and influential factors. A cross-sectional study.

Childhood obesity is a significant public health issue with far-reaching implications. The World Health Organization reported that in 2020, around 38 million children under five were overweight or obese globally, and in 2016, 340 million children and adolescents aged 5-19 were affected. In Pakistan, the situation is alarming; 66.9% of school-going children in Punjab were overweight, and 5.8% were obese in 2022. This study explores mothers' knowledge, attitudes, and practices regarding childhood obesity in Pakistan and identifies factors contributing to this epidemic. Maternal perspectives are crucial as they significantly influence children's dietary habits, physical activity, and attitudes toward food and body image. A cross-sectional study was conducted from November 2023 to January 2024 at four medical centers: Fazaia Ruth Pfau Medical College Hospitals Karachi, Baqai Medical University Karachi, and Nishtar Medical University Multan. The study included 191 mothers with children aged 5 to 15 years. Data were collected using a structured questionnaire on sociodemographic characteristics, knowledge, attitudes, practices, and perceptions of childhood obesity. IBM-SPSS version 26.0 was used for data analysis, employing statistical tests like Kruskal-Wallis, Mann-Whitney U, Spearman, or Kendall Tau correlation to examine associations. Among the 191 mothers, 48.7% had education beyond intermediate, and 86.9% were housewives. The children's BMI distribution showed that 27.7% were obese and 21.5% overweight. Mothers had moderate knowledge (60.5%) about childhood obesity; 75.4% recognized its long-term health risks, and 62.8% associated it with diabetes. Attitudes were generally positive, with a 78.5% average score. Most mothers (73.8%) believed obesity could be controlled and had healthy practices (70.1% average score). However, 96.9% reported witnessing stigmatization of obese children, and 79.6% felt pressured by relatives about their child's weight. The findings indicate that while Pakistani mothers have moderate knowledge and positive attitudes toward childhood obesity, their practices are influenced by educational and socioeconomic factors. Addressing these disparities, enhancing public health initiatives, and mitigating stigmatization could improve childhood obesity management in Pakistan.

AI based web application for diet planning and recipe generation

Introduces a web-based diet planning system based on machine learning, specifically Random Forests, deep learning and image processing capabilities to offer personalized nutritional input. Based on users' health data, their calorie needs, and food choices, the system is tailored to design diet plans specifically for each user. Finally, the system has functionalities of image processing capabilities: it identifies ingredients that feature in uploaded images of available food and thus comes out with recipes based on whatever is available. The goal of this integration is to make meal planning less painful, especially to individuals who have conditions like diabetes or hypertension, with respect to making healthier diets. Preliminary results demonstrate very good accuracy in recommending diets and identifying ingredients; these findings are positive toward the impact of the entire system in personalized nutrition management.

Do Social Media Likes Affect Food Consumption?

Social norms can affect food consumption. Although social media could be used to disseminate such norms, there is limited experimental research on the subject. The purpose of the study was to examine the effects of socially endorsed social media food posts, in the form of Instagram likes, on participants’ eating behavior of wheat food products. The survey was conducted in sensory booths where 149 participants were assigned to one of three conditions viewing three types of images: traditional Greek foods, modern foods, and home decoration as control. However, only one type was socially endorsed with likes. Participants self-reported on a series of questionnaires and were offered traditional Greek rusks (paximathia) and modern crackers as a snack reward during break time on two separate dishes. The hypothesis was that those who had paid attention to socially endorsed images of traditional foods would prefer to consumer Greek rusks, whereas those having paid attention to socially endorsed modern foods would opt for the crackers. Using the ANCOVA model, there was not identified any significant effect of condition on rusk or cracker consumption in grams after controlling the covariates. The results propose that exposure to socially endorsed images with Instagram likes, as a form of social media norm, do not incite people to consume more of either traditional rusks or modern crackers.

How does viewing food products in 2D or 3D affect consumer evaluations?

This study examines the influence of changes in visual information (2D vs. 3D) and product characteristics (healthy vs. unhealthy) on consumer purchase intention, expected satiety, and expected volume with regard to food products. The participants were asked to answer a questionnaire for each food sample displayed in 2D or 3D. The results reveal that the 3D food model has significantly higher scores for expected volume and satiety than does the 2D model, along with a significant interaction between visual information and product characteristics. For healthy foods, the expected volume, expected satiety, and purchase intention were significantly higher in the 2D than in the 3D model. For unhealthy foods, scores for the three variables were significantly higher in the 3D than in the 2D food model. This indicates an inverse relationship based on the healthiness of the food product. The study posits that changing the visual information from 2D to 3D strengthened the image of healthy and unhealthy foods as having low and high volumes, respectively, resulting in a more favorable purchase intention. The results of this study advance our understanding of the applicability of different image types when promoting food products, thereby enabling and guiding well-informed decisions of online retailers and other stakeholders in the design of food products for advertisements and other purposes in the food service industry.

Olfactory modulation of visual attention and preference towards congruent food products: An eye tracking study

In multisensory environments, odours often accompany visual stimuli, directing attention towards congruent objects. While previous research shows that people fixate longer on objects that match a recently smelled odour, it remains unclear whether odours directly influence product choices. Since odours persist in real-world settings, we investigated the effects of repeated odour exposure on visual attention and product choice, accounting for potential olfactory habituation. In a within-participant design, 30 participants completed a task where either a lemon odour (experimental condition) or clean air (control) was paired with congruent lemon-based food images, which varied to prevent visual habituation. We measured eye movements and choice preferences for these food products.Results revealed that participants exhibited longer gaze durations and more frequent fixations on food products congruent with the lemon odour. Repeated odour exposure had no effect on gaze patterns, as participants consistently focused on odour-congruent products throughout the experiment. The intensity and pleasantness of the lemon odour remained stable over time, suggesting no olfactory habituation occurred with this food-related odour. Despite this stable visual attention and odour intensity and pleasantness, participants began to diversify their product choices, selecting fewer odour-congruent items over time.These findings suggest that while odours continue to direct attention towards matching products, repeated exposure may reduce their influence on product choice, highlighting the complex role of olfactory stimuli in decision-making. The study provides insights into how odours interact with visual cues and influence consumer behaviour in prolonged exposure scenarios.

Lightweight Food Recognition via Aggregation Block and Feature Encoding

Food image recognition has recently been given considerable attention in the multimedia field in light of its possible implications on health. The characteristics of the dispersed distribution of ingredients in food images put forward higher requirements on the long-range information extraction ability of neural networks, leading to more complex and deeper models. Nevertheless, the lightweight version of food image recognition is essential for improved implementation on end devices and sustained server-side expansion. To address this issue, we present Aggregation Feature Net (AFNet), a lightweight network that is capable of effectively capturing both global and local features from food images. In AFNet, we develop a novel convolution based on a residual model by encoding global features through row-wise and column-wise information integration. Merging aggregation block with classic local convolution yields a framework that works as the backbone of the network. Based on the efficient use of parameters by the aggregation block, we constructed a lightweight food image recognition network with fewer layers and a smaller scale, assisted by a new type of activation function. Experimental results on four popular food recognition datasets demonstrate that our approach achieves state-of-the-art performance with higher accuracy and fewer FLOPs and parameters. For example, in comparison to the current state-of-the-art model of MobileViTv2, AFNet achieved 88.4% accuracy of the top-1 level on the ETHZ Food-101 dataset, with similar parameters and FLOPs but 1.4% more accuracy. The source code will be provided in supplementary materials.

The Effect of Food Image, Food Satisfaction on Culinary Quality and Behavioral Intention in Sampang Regency

This study aims to explore how perceptions of local food (food image) and consumer satisfaction with culinary experiences (food satisfaction) influence the perception of the culinary quality of the area and how these factors influence consumer behavior, such as intention to revisit or recommend. This study was conducted in Sampang Regency, an area known for its rich culinary heritage. The study population was local culinary consumers who had visited Pak Salim's Bebek Songkem restaurant located in Sampang Regency. Purposive sampling was used to determine respondents with a sample of 193 respondents. Structural equations were used as the analysis method using Smart PLS 4.0. Data were collected through a survey distributed to consumers who enjoyed Bebek Songkem. The study findings revealed that food image had a significant positive effect on culinary quality, while food satisfaction played an important role in strengthening this perception. In addition, food image and food satisfaction significantly influenced behavioral intentions, with higher culinary quality leading to a greater likelihood of positive consumer actions such as revisits and recommendations. These results underscore the importance of improving the image and satisfaction of local culinary experiences to promote Sampang as a culinary destination. This research offers practical insights for local businesses and policy makers in developing strategies to enhance culinary tourism and culinary experiences in the region.

Transfer Learning for Turkish Cuisine Classification

Thanks to developments in data-oriented domains like deep learning and big data, the integration of artificial intelligence with food category recognition has been a topic of interest for decades. The capacity of image classification to produce more precise outcomes in less time has made it a popular topic in computer vision. For the purpose of food categorization, three well-known CNN-based models—EfficientNetV2M, ResNet101, and VGG16—were fine-tuned in this research. Moreover, the pre-trained Vision Transformer (ViT) was used for feature extraction, followed by classification using a Random Forest (RF) algorithm. All the models were assessed on the TurkishFoods-15 dataset. It was found that the ViT and RF models were most effective in accurately capturing food images, with precision, recall, and F1-score values of 0.91, 0.86, and 0.88 respectively.

Enhancing Food Image Classification with Particle Swarm Optimization on NutriFoodNet and Data Augmentation Parameters

Enhancing Food Image Classification with Particle Swarm Optimization on NutriFoodNet and Data Augmentation Parameters