MMFood'25, the 1st International Workshop on Multi-modal Food Computing, brings together researchers and practitioners at the intersection of artificial intelligence, computer vision, natural language processing, and sensory modeling to advance the study of food. The workshop highlights how multimodal methods can be applied to food recognition, recommendation, analysis, and monitoring to address pressing challenges in health, nutrition, sustainability, and food culture. It features a rich program including a keynote, an invited talk, paper presentations, a poster session, and a panel discussion on the role of multimodal AI in preserving cultural heritage, fostering sustainable food futures, and enabling personal well-being. By convening experts from academia, industry, and healthcare, MMFood'25 provides a unique platform for fostering interdisciplinary collaboration and for shaping the emerging field of multimodal food computing. The workshop proceedings can be found at: https://dl.acm.org/doi/proceedings/10.1145/3746264.

The advancement of artificial intelligence (AI) in food and nutrition research is hindered by a critical bottleneck: the lack of annotated food data. Despite the rise of highly efficient AI models designed for tasks such as food segmentation and classification, their practical application might necessitate proficiency in AI and machine learning principles, which can act as a challenge for non-AI experts in the field of nutritional sciences. Alternatively, it highlights the need to translate AI models into user-friendly tools that are accessible to all. To address this, we present a demo of a semi-automatic food image annotation tool leveraging the Segment Anything Model (SAM) kirillov2023segment??>. The tool enables prompt-based food segmentation via user interactions, promoting user engagement and allowing them to further categorise food items within meal images and specify weight/volume if necessary. Additionally, we release a fine-tuned version of SAM’s mask decoder, dubbed MealSAM, with the ViT-B backbone tailored specifically for food image segmentation. Our objective is not only to contribute to the field by encouraging participation, collaboration, and the gathering of more annotated food data but also to make AI technology available for a broader audience by translating AI into practical tools.

This abstract provides a summary and overview of the 8th International Workshop on Multimedia Assisted Dietary Management.

This abstract provides a summary and overview of the 7th International Workshop on Multimedia Assisted Dietary Management.

Regular monitoring of nutrient intake in hospitalised patients plays a critical role in reducing the risk of disease-related malnutrition. Although several methods to estimate nutrient intake have been developed, there is still a clear demand for a more reliable and fully automated technique, as this could improve data accuracy and reduce both the burden on participants and health costs. In this paper, we propose a novel system based on artificial intelligence (AI) to accurately estimate nutrient intake, by simply processing RGB Depth (RGB-D) image pairs captured before and after meal consumption. The system includes a novel multi-task contextual network for food segmentation, a few-shot learning-based classifier built by limited training samples for food recognition, and an algorithm for 3D surface construction. This allows sequential food segmentation, recognition, and estimation of the consumed food volume, permitting fully automatic estimation of the nutrient intake for each meal. For the development and evaluation of the system, a dedicated new database containing images and nutrient recipes of 322 meals is assembled, coupled to data annotation using innovative strategies. Experimental results demonstrate that the estimated nutrient intake is highly correlated (>0.91) to the ground truth and shows very small mean relative errors (<20%), outperforming existing techniques proposed for nutrient intake assessment.

Food volume estimation is an essential step in the pipeline of dietary assessment and demands the precise depth estimation of the food surface and table plane. Existing methods based on computer vision require either multi-image input or additional depth maps, reducing convenience of implementation and practical significance. Despite the recent advances in unsupervised depth estimation from a single image, the achieved performance in the case of large texture-less areas needs to be improved. In this paper, we propose a network architecture that jointly performs geometric understanding (i.e., depth prediction and 3D plane estimation) and semantic prediction on a single food image, enabling a robust and accurate food volume estimation regardless of the texture characteristics of the target plane. For the training of the network, only monocular videos with semantic ground truth are required, while the depth map and 3D plane ground truth are no longer needed. Experimental results on two separate food image databases demonstrate that our method performs robustly on texture-less scenarios and is superior to unsupervised networks and structure from motion based approaches, while it achieves comparable performance to fully-supervised methods.

Image registration in multimodal, multitemporal satellite imagery is one of the most important problems in remote sensing and essential for a number of other tasks such as change detection and image fusion. In this paper, inspired by the recent success of deep learning approaches we propose a novel convolutional neural network architecture that couples linear and deformable approaches for accurate alignment of remote sensing imagery. The proposed method is completely unsupervised, ensures smooth displacement fields and provides real time registration on a pair of images. We evaluate the performance of our method using a challenging multitemporal dataset of very high resolution satellite images and compare its performance with a state of the art elastic registration method based on graphical models. Both quantitative and qualitative results prove the high potentials of our method.

In this study, we propose a 3D deep neural network called U-ReSNet, a joint framework that can accurately register and segment medical volumes. The proposed network learns to automatically generate linear and elastic deformation models, trained by minimizing the mean square error and the local cross correlation similarity metrics. In parallel, a coupled architecture is integrated, seeking to provide segmentation maps for anatomies or tissue patterns using an additional decoder part trained with the dice coefficient metric. U-ReSNet is trained in an end to end fashion, while due to this joint optimization the generated network features are more informative leading to promising results compared to other deep learning-based methods existing in the literature. We evaluated the proposed architecture using the publicly available OASIS 3 dataset, measuring the dice coefficient metric for both registration and segmentation tasks. Our promising results indicate the potentials of our method which is composed from a convolutional architecture that is extremely simple and light in terms of parameters.

This abstract provides a summary and overview of the 5th International Workshop on Multimedia Assisted Dietary Management.

Early and accurate diagnosis of interstitial lung diseases (ILDs) is crucial for making treatment decisions, but can be challenging even for experienced radiologists. The diagnostic procedure is based on the detection and recognition of the different ILD pathologies in thoracic CT scans, yet their manifestation often appears similar. In this study, we propose the use of a deep purely convolutional neural network for the semantic segmentation of ILD patterns, as the basic component of a computer aided diagnosis system for ILDs. The proposed CNN, which consists of convolutional layers with dilated filters, takes as input a lung CT image of arbitrary size and outputs the corresponding label map. We trained and tested the network on a data set of 172 sparsely annotated CT scans, within a cross-validation scheme. The training was performed in an end-to-end and semisupervised fashion, utilizing both labeled and nonlabeled image regions. The experimental results show significant performance improvement with respect to the state of the art.

Self-monitoring of blood glucose (SMBG) and continuous glucose monitoring (CGM) are commonly used by type 1 diabetes (T1D) patients to measure glucose concentrations. The proposed adaptive basal-bolus algorithm (ABBA) supports inputs from either SMBG or CGM devices to provide personalised suggestions for the daily basal rate and prandial insulin doses on the basis of the patients' glucose level on the previous day. The ABBA is based on reinforcement learning, a type of artificial intelligence, and was validated in silico with an FDA-accepted population of 100 adults under different realistic scenarios lasting three simulated months. The scenarios involve three main meals and one bedtime snack per day, along with different variabilities and uncertainties for insulin sensitivity, mealtime, carbohydrate amount, and glucose measurement time. The results indicate that the proposed approach achieves comparable performance with CGM or SMBG as input signals, without influencing the total daily insulin dose. The results are a promising indication that AI algorithmic approaches can provide personalised adaptive insulin optimization and achieve glucose control-independent of the type of glucose monitoring technology.

An effective surveillance strategy for the progression of abdominal aortic aneurysms (AAAs) may be achieved by assessing its expected growth rate in a personalized manner. Given the variety of factors with an impact on AAA growth, an integrative approach to the problem could potentially benefit from incorporating clinical and morphometric data, as well as mechanical stress characterizations. In addition, here we investigated the use of texture information on computed tomography angiography images within the AAA sac. A cohort of n=38 patients underwent a baseline examination, plus a follow-up visit to measure AAA growth rates, in terms of its maximum diameter (Dmax) divided by the elapsed time period. Subsequently, each case was labelled as slow, medium or quick growth, compared to the expected rate reported in demographic studies, as a function of gender and baseline Dmax. We computed a total of 102 features (5 clinical, 17 morphometric, 4 biomechanical, and 76 on texture) and used a number of machine learning (ML) algorithms; with the aim of minimizing misclassification costs. The performance of the system was evaluated with a leave-one-out cross-validation scheme. The results achieved by the best performing approach, an ensemble of decision trees ('LPBoost') using the entire 102-dimensional feature space, indicated that the combination of different information sources, along with ML algorithms, may have a positive impact on the AAA prognosis assessment.

Early diagnosis of interstitial lung diseases is crucial for their treatment, but even experienced physicians find it difficult, as their clinical manifestations are similar. In order to assist with the diagnosis, computer-aided diagnosis systems have been developed. These commonly rely on a fixed scale classifier that scans CT images, recognizes textural lung patterns, and generates a map of pathologies. In a previous study, we proposed a method for classifying lung tissue patterns using a deep convolutional neural network (CNN), with an architecture designed for the specific problem. In this study, we present an improved method for training the proposed network by transferring knowledge from the similar domain of general texture classification. Six publicly available texture databases are used to pretrain networks with the proposed architecture, which are then fine-tuned on the lung tissue data. The resulting CNNs are combined in an ensemble and their fused knowledge is compressed back to a network with the original architecture. The proposed approach resulted in an absolute increase of about 2% in the performance of the proposed CNN. The results demonstrate the potential of transfer learning in the field of medical image analysis, indicate the textural nature of the problem and show that the method used for training a network can be as important as designing its architecture.

The papers in this special section address the concept of nutrition informatics from food monitoring to dietary management. Non-communicable diseases (NCD) account for a massively increasing proportion of the global health burden. A number of behavioral and physiological factors are related to the rising onset of NCD worldwide, with unhealthy eating playing a key role among them. In parallel, food allergies and associated acute and sometimes life-threatening reactions are a public health problem. Thus, balanced nutrition with a proper diet is the key to the prevention of diet related diseases. The recent advances in smartphone technologies, wearable sensors, computer vision and machine learning will bring the applications of nutrition informatics closer to the individuals and enable them to make better decisions regarding their daily lives.

The increasing prevalence of diet-related chronic diseases coupled with the ineffectiveness of traditional diet management methods have resulted in a need for novel tools to accurately and automatically assess meals. Recently, computer vision-based systems that use meal images to assess their content have been proposed. Food portion estimation is the most difficult task for individuals assessing their meals and it is also the least studied area. This paper proposes a three-stage system to calculate portion sizes using two images of a dish acquired by mobile devices. The first stage consists in understanding the configuration of the different views, after which a dense three-dimensional (3D) model is built from the two images; finally, this 3D model serves to extract the volume of the different items. The system was extensively tested on 77 real dishes of known volume, and achieved an average error of less than 10% in 5.5 seconds per dish. The proposed pipeline is computationally tractable and requires no user input, making it a viable option for fully automated dietary assessment.

This abstract provides a summary and overview of the 2nd international workshop on multimedia assisted dietary management.

Automated tissue characterization is one of the most crucial components of a computer aided diagnosis (CAD) system for interstitial lung diseases (ILDs). Although much research has been conducted in this field, the problem remains challenging. Deep learning techniques have recently achieved impressive results in a variety of computer vision problems, raising expectations that they might be applied in other domains, such as medical image analysis. In this paper, we propose and evaluate a convolutional neural network (CNN), designed for the classification of ILD patterns. The proposed network consists of 5 convolutional layers with 2 × 2 kernels and LeakyReLU activations, followed by average pooling with size equal to the size of the final feature maps and three dense layers. The last dense layer has 7 outputs, equivalent to the classes considered: healthy, ground glass opacity (GGO), micronodules, consolidation, reticulation, honeycombing and a combination of GGO/reticulation. To train and evaluate the CNN, we used a dataset of 14696 image patches, derived by 120 CT scans from different scanners and hospitals. To the best of our knowledge, this is the first deep CNN designed for the specific problem. A comparative analysis proved the effectiveness of the proposed CNN against previous methods in a challenging dataset. The classification performance ( ~ 85.5%) demonstrated the potential of CNNs in analyzing lung patterns. Future work includes, extending the CNN to three-dimensional data provided by CT volume scans and integrating the proposed method into a CAD system that aims to provide differential diagnosis for ILDs as a supportive tool for radiologists.

Computer vision-based food recognition could be used to estimate a meal's carbohydrate content for diabetic patients. This study proposes a methodology for automatic food recognition, based on the bag-of-features (BoF) model. An extensive technical investigation was conducted for the identification and optimization of the best performing components involved in the BoF architecture, as well as the estimation of the corresponding parameters. For the design and evaluation of the prototype system, a visual dataset with nearly 5000 food images was created and organized into 11 classes. The optimized system computes dense local features, using the scale-invariant feature transform on the HSV color space, builds a visual dictionary of 10000 visual words by using the hierarchical k-means clustering and finally classifies the food images with a linear support vector machine classifier. The system achieved classification accuracy of the order of 78%, thus proving the feasibility of the proposed approach in a very challenging image dataset.

In this paper, we propose novel methodologies for the automatic segmentation and recognition of multi-food images. The proposed methods implement the first modules of a carbohydrate counting and insulin advisory system for type 1 diabetic patients. Initially the plate is segmented using pyramidal mean-shift filtering and a region growing algorithm. Then each of the resulted segments is described by both color and texture features and classified by a support vector machine into one of six different major food classes. Finally, a modified version of the Huang and Dom evaluation index was proposed, addressing the particular needs of the food segmentation problem. The experimental results prove the effectiveness of the proposed method achieving a segmentation accuracy of 88.5% and recognition rate equal to 87%.

There is great demand for easily-accessible, user-friendly dietary self-management applications. Yet accurate, fully-automatic estimation of nutritional intake using computer vision methods remains an open research problem. One key element of this problem is the volume estimation, which can be computed from 3D models obtained using multi-view geometry. The paper presents a computational system for volume estimation based on the processing of two meal images. A 3D model of the served meal is reconstructed using the acquired images and the volume is computed from the shape. The algorithm was tested on food models (dummy foods) with known volume and on real served food. Volume accuracy was in the order of 90 %, while the total execution time was below 15 seconds per image pair. The proposed system combines simple and computational affordable methods for 3D reconstruction, remained stable throughout the experiments, operates in near real time, and places minimum constraints on users.

Chronic care of diabetes comes with large amounts of data concerning the self- and clinical management of the disease. In this paper, we propose to treat that information from two different perspectives. Firstly, a predictive model of short-term glucose homeostasis relying on machine learning is presented with the aim of preventing hypoglycemic events and prolonged hyperglycemia on a daily basis. Second, data mining approaches are proposed as a tool for explaining and predicting the long-term glucose control and the incidence of diabetic complications.

BACKGROUND: Obesity is a multifactorial trait, which comprises an independent risk factor for cardiovascular disease (CVD). The aim of the current work is to study the complex etiology beneath obesity and identify genetic variations and/or factors related to nutrition that contribute to its variability. To this end, a set of more than 2300 white subjects who participated in a nutrigenetics study was used. For each subject a total of 63 factors describing genetic variants related to CVD (24 in total), gender, and nutrition (38 in total), e.g. average daily intake in calories and cholesterol, were measured. Each subject was categorized according to body mass index (BMI) as normal (BMI ≤ 25) or overweight (BMI > 25). Two artificial neural network (ANN) based methods were designed and used towards the analysis of the available data. These corresponded to i) a multi-layer feed-forward ANN combined with a parameter decreasing method (PDM-ANN), and ii) a multi-layer feed-forward ANN trained by a hybrid method (GA-ANN) which combines genetic algorithms and the popular back-propagation training algorithm. RESULTS: PDM-ANN and GA-ANN were comparatively assessed in terms of their ability to identify the most important factors among the initial 63 variables describing genetic variations, nutrition and gender, able to classify a subject into one of the BMI related classes: normal and overweight. The methods were designed and evaluated using appropriate training and testing sets provided by 3-fold Cross Validation (3-CV) resampling. Classification accuracy, sensitivity, specificity and area under receiver operating characteristics curve were utilized to evaluate the resulted predictive ANN models. The most parsimonious set of factors was obtained by the GA-ANN method and included gender, six genetic variations and 18 nutrition-related variables. The corresponding predictive model was characterized by a mean accuracy equal of 61.46% in the 3-CV testing sets. CONCLUSIONS: The ANN based methods revealed factors that interactively contribute to obesity trait and provided predictive models with a promising generalization ability. In general, results showed that ANNs and their hybrids can provide useful tools for the study of complex traits in the context of nutrigenetics.

SMARTDIAB is a platform designed to support the monitoring, management, and treatment of patients with type 1 diabetes mellitus (T1DM), by combining state-of-the-art approaches in the fields of database (DB) technologies, communications, simulation algorithms, and data mining. SMARTDIAB consists mainly of two units: 1) the patient unit (PU); and 2) the patient management unit (PMU), which communicate with each other for data exchange. The PMU can be accessed by the PU through the internet using devices, such as PCs/laptops with direct internet access or mobile phones via a Wi-Fi/General Packet Radio Service access network. The PU consists of an insulin pump for subcutaneous insulin infusion to the patient and a continuous glucose measurement system. The aforementioned devices running a user-friendly application gather patient's related information and transmit it to the PMU. The PMU consists of a diabetes data management system (DDMS), a decision support system (DSS) that provides risk assessment for long-term diabetes complications, and an insulin infusion advisory system (IIAS), which reside on a Web server. The DDMS can be accessed from both medical personnel and patients, with appropriate security access rights and front-end interfaces. The DDMS, apart from being used for data storage/retrieval, provides also advanced tools for the intelligent processing of the patient's data, supporting the physician in decision making, regarding the patient's treatment. The IIAS is used to close the loop between the insulin pump and the continuous glucose monitoring system, by providing the pump with the appropriate insulin infusion rate in order to keep the patient's glucose levels within predefined limits. The pilot version of the SMARTDIAB has already been implemented, while the platform's evaluation in clinical environment is being in progress.

Identification of gene-gene and gene-environment interactions that contribute in the onset of a multi-factorial disease supports the prevention of diseases like the Cardiovascular Disease (CVD). Body Mass Index (BMI), a measure of human obesity, is an independent risk factor of CVD. Furthermore, it is known that a subjectpsilas BMI is affected both by his/her lifestyle, e.g. nutrition, and genetic profile. Aim of the paper is to predict a subjectpsilas onset of obesity using lifestyle and genetic information. The prediction is performed by a computational intelligence based system using a Parameter Decreasing Method (PDM) combined with an Artificial Neural Network (ANN). The system uses an initial set of 63 input variables corresponding to sex, average nutrition intake measurements, and genetic variations to identify the 32 most important ones that affect BMI. The selected variables are the ones to interact with each other towards the complex trait of BMI, which is used as a 2-class output variable (BMI les 25 vs. BMIges25) in the ANN. The system achieved a mean accuracy of the system evaluated by a 3-cross validation resampling technique equal to 77.89%.

A computer aided diagnosis (CAD) system for the characterization of hepatic tissue from computed tomography (CT) images is presented. Regions of interest (ROI's) corresponding to four types of hepatic tissue are drawn by an experienced radiologist on abdominal non-enhanced CT images. For each ROI, five sets of texture features are extracted and combined to provide input to the CAD system. If the dimensionality of a feature set is greater than a predefined threshold, appropriate feature selection based on a genetic algorithm (GA) is applied. Classification of the ROI is then carried out using an ensemble of classifiers consisting of two neural network (NN) and three statistical classifiers. The final decision of the CAD system is based on the application of a voting scheme across the outputs of the primary classifiers of the ensemble. A classification performance of the order of 90.63% was finally achieved.

In this paper, a computer-aided diagnostic (CAD) system for the classification of hepatic lesions from computed tomography (CT) images is presented. Regions of interest (ROIs) taken from nonenhanced CT images of normal liver, hepatic cysts, hemangiomas, and hepatocellular carcinomas have been used as input to the system. The proposed system consists of two modules: the feature extraction and the classification modules. The feature extraction module calculates the average gray level and 48 texture characteristics, which are derived from the spatial gray-level co-occurrence matrices, obtained from the ROIs. The classifier module consists of three sequentially placed feed-forward neural networks (NNs). The first NN classifies into normal or pathological liver regions. The pathological liver regions are characterized by the second NN as cyst or "other disease." The third NN classifies "other disease" into hemangioma or hepatocellular carcinoma. Three feature selection techniques have been applied to each individual NN: the sequential forward selection, the sequential floating forward selection, and a genetic algorithm for feature selection. The comparative study of the above dimensionality reduction methods shows that genetic algorithms result in lower dimension feature vectors and improved classification performance.