Early detection of chronic and non-communicable diseases (NCDs) is crucial for effective treatment during the initial stages. This study explores the application of wearable devices and artificial intelligence (AI) to predict weight loss changes in overweight and obese individuals. To the best of our knowledge, this is the first study to leverage wearable-derived physiological and behavioral features for weight loss prediction using AI. Using wearable data from a 1-month trial involving around 100 subjects from the AI4FoodDB database, including biomarkers, vital signs, and behavioral data, we identify key differences between those achieving weight loss (≥ 2% of their initial weight) and those who do not. Feature selection techniques highlight key predictors, including heart rate variability, sleep duration, and physical activity patterns, underscoring the interplay between physiological and behavioral factors. Classification algorithms reveal promising results, with the Gradient Boosting classifier achieving 84.44% area under the curve (AUC). The integration of multiple data sources (e.g., vital signs, physical and sleep activity, etc.) enhances performance, suggesting the potential of wearable devices and AI in personalized healthcare.

Recent advancements in Vision-Language Models (VLMs) have demonstrated strong capabilities in general visual reasoning, yet their applicability to rigorous biometric tasks remains unexplored. This work presents an exploratory study evaluating the zero-shot performance of state-of-the-art VLMs (GPT-5.2 and Gemini 2.5 Pro) on the Signature Verification Challenge (SVC) benchmark. To enable visual processing, raw kinematic time-series are converted into static images, encoding pressure information into stroke opacity whenever available in the source data. Furthermore, we introduce a scoring protocol that extracts latent token probabilities to compute robust biometric scores. Experimental results reveal a significant performance dichotomy dependent on signal quality and forgery type. In random forgery scenarios, the zero-shot VLM achieves exceptional discrimination, with GPT-5.2 reaching an Equal Error Rate of 0.32

Document Visual Question Answering (DocVQA) presents a complex multimodal challenge, requiring models to exploit visual, textual, and layout information from documents. Although Vision-Language Models (VLMs) have shown remarkable performance in text-vision tasks, their robustness and transferability to different document domains remains underexplored. In this study, we present a comprehensive evaluation of 8 open-source pretrained VLMs on DocVQA in three different document domains: industrial documents of varying type, infographics, and presentation slides. We systematically assess model performance under zero-shot evaluations, fully supervised finetuning with inter- and intra-dataset evaluations, and few-shot learning evaluations of knowledge transfer between domains. Our findings demonstrate that while large pretrained VLMs possess strong zero-shot baselines for structured layouts, their performance strongly decreases on visually complex layouts of infographics and slides. Although parameter scaling is a dominant factor on performance, supervised finetuning yields higher relative gains in smaller architectures. Furthermore, our cross-domain and few-shot experiments show that visual understanding is the main bottleneck for DocVQA, not a lack of knowledge from the VLMs. Using 50 target domain samples, the models finetuned in DocVQA with datasets of different domains rapidly adapt to the target domain documents, even surpassing their fully supervised counterparts in some cases.

This article presents the Keystroke Verification Challenge- onGoing (KVC-onGoing)1, on which researchers can easily benchmark their systems in a common platform using large-scale public databases, the Aalto University Keystroke databases, and a standard experimental protocol. The keystroke data consist of tweet- long sequences of variable transcript text from over 185,000 subjects, acquired through desktop and mobile keyboards simulating real-life conditions. The results on the evaluation set of KVC-onGoing have proved the high discriminative power of keystroke dynamics, reaching values as low as 3.33% of Equal Error Rate (EER) and 11.96% of False Non-Match Rate (FNMR) @1% False Match Rate (FMR) in the desktop scenario, and 3.61% of EER and 17.44% of FNMR @1% at FMR in the mobile scenario, significantly improving previous state-of-the-art results. Concerning demographic fairness, the analyzed scores reflect the subjects' age and gender to various extents, not negligible in a few cases. The framework runs on CodaLab2.

BACKGROUND & AIMS:The widespread prevalence of NCDs calls for an improvement in their prevention and treatment. Wearable technologies can be an important asset in the development of precision nutrition strategies, for both health professionals and patients. However, their clinical use is hindered by a lack of validation against current methodologies or appropriate tools to deliver nutritional strategies based on their data. Our study includes manual and automatic data capture methods within a weight loss intervention with the aim to create an essential asset for the implementation, validation, and benchmarking of AI-based tools in nutritional clinical practice. METHODS:This is a feasibility prospective and crossover controlled trial for weight loss in overweight and obese participants, randomized into two groups: Group 1 used manual data collection methods based on validated questionnaires for the first two weeks; while Group 2 started with automatic data collection methods consisting of wearable sensors. After two weeks, the two groups switched data collection methods. Lifestyle data, anthropometric measurements and biological samples were collected from all participants. RESULTS:A total of 93 participants completed the nutritional intervention designed for weight loss, achieving a mean reduction of 2 kg (V1: 84.99 SD ± 13.69, V3: 82.72 SD ± 13.32, p < 0.001). Significant reductions were observed in body mass index, visceral fat, waist circumference, total cholesterol, and HbA1c levels. The use of electronic devices proved satisfactory among the participants (System Usability Scale score 78.27 ± 12.86). We also report the presence of distinct patient groups based on continuous glucose measurements. CONCLUSION:This study has yielded a large amount of data and has showcased how automatic data collection devices can be employed to gather data in the context of a nutritional intervention. This will enable the implementation of AI-based tools in nutritional clinical practice. CLINICAL TRIAL REGISTRATION:ClinicalTrials.gov, NCT05807243.

Synthetic data is gaining increasing popularity for face recognition technologies, mainly due to the privacy concerns and challenges associated with obtaining real data, including diverse scenarios, quality, and demographic groups, among others. It also offers some advantages over real data, such as the large amount of data that can be generated or the ability to customize it to adapt to specific problem-solving needs. To effectively use such data, face recognition models should also be specifically designed to exploit synthetic data to its fullest potential. In order to promote the proposal of novel Generative AI methods and synthetic data, and investigate the application of synthetic data to better train face recognition systems, we introduce the 2nd FRCSyn-onGoing challenge, based on the 2nd Face Recognition Challenge in the Era of Synthetic Data (FRCSyn), originally launched at CVPR 2024. This is an ongoing challenge that provides researchers with an accessible platform to benchmark i) the proposal of novel Generative AI methods and synthetic data, and ii) novel face recognition systems that are specifically proposed to take advantage of synthetic data. We focus on exploring the use of synthetic data both individually and in combination with real data to solve current challenges in face recognition such as demographic bias, domain adaptation, and performance constraints in demanding situations, such as age disparities between training and testing, changes in the pose, or occlusions. Very interesting findings are obtained in this second edition, including a direct comparison with the first one, in which synthetic databases were restricted to DCFace and GANDiffFace.

Automatic dietary assessment based on food images remains a challenge, requiring precise food detection, segmentation, and classification. Vision-Language Models (VLMs) offer new possibilities by integrating visual and textual reasoning. In this study, we evaluate six state-of-the-art VLMs (ChatGPT, Gemini, Claude, Moondream, DeepSeek, and LLaVA), analyzing their capabilities in food recognition at different levels. For the experimental framework, we introduce the FoodNExTDB, a unique food image database that contains 9,263 expert-labeled images across 10 categories (e.g., "protein source"), 62 subcategories (e.g., "poultry"), and 9 cooking styles (e.g., "grilled"). In total, FoodNExTDB includes 50k nutritional labels generated by seven experts who manually annotated all images in the database. Also, we propose a novel evaluation metric, Expert-Weighted Recall (EWR), that accounts for the inter-annotator variability. Results show that closed-source models outperform open-source ones, achieving over 90

Behavioral biometrics based on smartphone motion sensors are growing in popularity for authentication purposes. In this study, AirSignatureDB is presented: a new publicly accessible dataset of in-air signatures collected from 108 participants under real-world conditions, using 83 different smartphone models across four sessions. This dataset includes genuine samples and skilled forgeries, enabling a comprehensive evaluation of system robustness against realistic attack scenarios. Traditional and deep learning-based methods for in-air signature verification are benchmarked, while analyzing the influence of sensor modality and enrollment strategies. Beyond verification, a first approach to reconstructing the three-dimensional trajectory of in-air signatures from inertial sensor data alone is introduced. Using on-line handwritten signatures as a reference, we demonstrate that the recovery of accurate trajectories is feasible, challenging the long-held assumption that in-air gestures are inherently traceless. Although this approach enables forensic traceability, it also raises critical questions about the privacy boundaries of behavioral biometrics. Our findings underscore the need for a reevaluation of the privacy assumptions surrounding inertial sensor data, as they can reveal user-specific information that had not previously been considered in the design of in-air signature systems.

This article introduces the Membership Inference Test (MINT), a novel approach that aims to empirically assess if given data was used during the training of AI/ML models. Specifically, we propose two MINT architectures designed to learn the distinct activation patterns that emerge when an Audited Model is exposed to data used during its training process. These architectures are based on Multilayer Perceptrons (MLPs) and Convolutional Neural Networks (CNNs). The experimental framework focuses on the challenging task of Face Recognition, considering three state-of-the-art Face Recognition systems. Experiments are carried out using six publicly available databases, comprising over 22 million face images in total. Different experimental scenarios are considered depending on the context of the AI model to test. Our proposed MINT approach achieves promising results, with up to 90% accuracy, indicating the potential to recognize if an AI model has been trained with specific data. The proposed MINT approach can serve to enforce privacy and fairness in several AI applications, e.g., revealing if sensitive or private data was used for training or tuning Large Language Models (LLMs).

Person re-identification (ReID) across aerial and ground vantage points has become crucial for large-scale surveillance and public safety applications. Although significant progress has been made in ground-only scenarios, bridging the aerial-ground domain gap remains a formidable challenge due to extreme viewpoint differences, scale variations, and occlusions. Building upon the achievements of the AG-ReID 2023 Challenge, this paper introduces the AG-VPReID 2025 Challenge - the first large-scale video-based competition focused on high-altitude (80-120m) aerial-ground ReID. Constructed on the new AG-VPReID dataset with 3,027 identities, over 13,500 tracklets, and approximately 3.7 million frames captured from UAVs, CCTV, and wearable cameras, the challenge featured four international teams. These teams developed solutions ranging from multi-stream architectures to transformer-based temporal reasoning and physics-informed modeling. The leading approach, X-TFCLIP from UAM, attained 72.28

Active Membership Inference Test (aMINT) is a method designed to detect whether given data were used during the training of machine learning models. In Active MINT, we propose a novel multitask learning process that involves training simultaneously two models: the original or Audited Model, and a secondary model, referred to as the MINT Model, responsible for identifying the data used for training the Audited Model. This novel multi-task learning approach has been designed to incorporate the auditability of the model as an optimization objective during the training process of neural networks. The proposed approach incorporates intermediate activation maps as inputs to the MINT layers, which are trained to enhance the detection of training data. We present results using a wide range of neural networks, from lighter architectures such as MobileNet to more complex ones such as Vision Transformers, evaluated in 5 public benchmarks. Our proposed Active MINT achieves over 80

In the digital era, the amount of digital documents generated each day have being increasing exponentially with the years, to a point where it is unfeasible to process them manually. Thus, there has been growing interest from different sectors to develop automatic tools to process digital documents in an automatic manner. Yet useful, this task is challenging, due to both the large variability and the multimodal nature inherent to the problem. In most cases, a text-only approach often falls short in comprehending the information conveyed by diverse components of varying significance. In this regard, Document Layout Analysis (DLA) has been an interesting research field for many years, whose objective it to detect and classify the basic components of a document. Thus, is an interesting task to obtain a first understanding on how the different components of the document interact with each other. In this work, we used a semi-automatic procedure to annotate digital documents with different layout labels, including 4 basic layout blocks and 4 text categories. We apply this procedure to collect a novel database for DLA in the public affairs domain, the PALdb database, using a set of 24 data sources from the Spanish Administration. The database comprises 37.9K documents with more than 441K document pages, and more than 8M labels associated to 8 layout block units. The results of our experiments validate the proposed text labeling procedure with accuracy up to 99%. We also present a novel application of Quickest Change Detection (QCD) techniques on the DLA domain, which we use to continuously detect changes in the layout of the documents from multiple sources.

This article provides a comprehensive overview of recent research in the area of Child-Computer Interaction (CCI). The main contributions of the present article are two-fold. First, we present a novel longitudinal CCI database named ChildCIdbLong, which comprises over 600 children aged 18 months to 8 years old, acquired continuously over 4 academic years (2019-2023). As a result, ChildCIdbLong comprises over 12K test acquisitions over a tablet device. Different tests are considered in ChildCIdbLong, requiring different touch and stylus gestures, enabling the evaluation of praxical and cognitive skills such as attentional, visuo-spatial, and executive, among others. In addition to the ChildCIdbLong database, we propose a novel quantitative metric called Test Quality (Q), designed to measure the motor and cognitive development of children through their interaction with a tablet device. In order to provide a better comprehension of the proposed Q metric, popular percentile-based growth representations are introduced for each test, providing a two-dimensional space to compare children's development with respect to the typical age skills of the population. The results achieved in the present article highlight the potential of the novel ChildCIdbLong database in conjunction with the proposed Q metric to measure the motor and cognitive development of children as they grow up. The proposed framework could be very useful as an automatic tool to support child experts (e.g., paediatricians, educators, or neurologists) for early detection of potential physical/cognitive impairments during children's development.

Obesity is a prevalent and preventable condition that has increased alarmingly in recent decades and is associated with low-grade chronic inflammation and a range of non-communicable diseases. Although the aetiology of obesity has been simplified as an energy imbalance, multiple factors contribute to its development, making its treatment complex. Current methods for diagnosing obesity and overweight have significant limitations therefore, a more personalised approach to nutritional guidance is needed to account for differences in response to dietary affected by environmental and genetic factors. In addition, nutritional therapy and research continue to face issues such as measurement errors, memory biases, and others. The objective of this work is to present the AI4Food project. Massive data capture has been carried out using traditional and digital methods throughout a nutritional intervention for healthy weight loss in order to compare both data collection methodologies and integrate all the information of the variables obtained with artificial intelligence. A total of 93 (28 males, 65 females, mean age of 49.60 (SD ± 12.78) years) participants completed one-month prospective and crossover nutritional intervention with a traditional and digital group. Anthropometric measurements, vital signs, clinical history, and lifestyle habits were registered in all visits. Moreover, in the traditional intervention the questionnaires: 3-day food record, physical activity (IPAQ), sleep pattern (COS), psychological and emotional state (DASS-21), and health status and quality of life (SF-36) were used. In the digital intervention, the sensors Freestyle Libre 2 Glucometer and smartwatch Fitbit Sense were used to monitor glucose levels and lifestyle (heart rate, heart rate variability, sleep patterns, electrodermal activity, etc.) and diet was monitored by image captures. Faecal and blood samples were collected for future analysis. Preliminary results are promising as both data collection methodologies have been successfully compared and all information on the variables obtained has been integrated. It will allow for improving and personalising the lifestyle recommendations given to the population.

Synthetic data is gaining increasing relevance for training machine learning models. This is mainly motivated due to several factors such as the lack of real data and intra-class variability, time and errors produced in manual labeling, and in some cases privacy concerns, among others. This paper presents an overview of the 2nd edition of the Face Recognition Challenge in the Era of Synthetic Data (FRCSyn) organized at CVPR 2024. FRCSyn aims to investigate the use of synthetic data in face recognition to address current technological limitations, including data privacy concerns, demographic biases, generalization to novel scenarios, and performance constraints in challenging situations such as aging, pose variations, and occlusions. Unlike the 1st edition, in which synthetic data from DCFace and GANDiffFace methods was only allowed to train face recognition systems, in this 2nd edition we propose new sub-tasks that allow participants to explore novel face generative methods. The outcomes of the 2nd FRCSyn Challenge, along with the proposed experimental protocol and benchmarking contribute significantly to the application of synthetic data to face recognition.

We analyze various factors affecting the proper functioning of MIA and MINT, two research lines aimed at detecting data used for training. The difference between these lines lies in the environmental conditions, while the fundamental bases are similar for both. As evident in the literature, this detection task is far from straightforward and poses an ongoing challenge for the scientific community. Specifically, in this work, we conclude that factors such as the number of times data passes through the original network, the loss function, or dropout significantly impact detection outcomes. Therefore, it is crucial to consider them when developing these methods and during the training of any neural network, both to avoid (MIA) and to enhance (MINT) this detection. We evaluate the AdaFace facial recognition model using five databases with over 22 million images, modifying the different factors under analysis and defining a suitable protocol for their examination. State-of-the-art accuracy reaching up to 87% is achieved, surpassing existing methods.

Despite the widespread adoption of face recognition technology around the world, and its remarkable performance on current benchmarks, there are still several challenges that must be covered in more detail. This paper offers an overview of the Face Recognition Challenge in the Era of Synthetic Data (FRCSyn) organized at WACV 2024. This is the first international challenge aiming to explore the use of synthetic data in face recognition to address existing limitations in the technology. Specifically, the FRCSyn Challenge targets concerns related to data privacy issues, demographic biases, generalization to unseen scenarios, and performance limitations in challenging scenarios, including significant age disparities between enrollment and testing, pose variations, and occlusions. The results achieved in the FRCSyn Challenge, together with the proposed benchmark, contribute significantly to the application of synthetic data to improve face recognition technology.

This article proposes a novel children-computer interaction (CCI) approach for the task of age group detection. This approach focuses on the automatic analysis of the time series generated from the interaction of the children with mobile devices. In particular, we extract a set of 25 time series related to spatial, pressure, and kinematic information of the children interaction while colouring a tree through a pen stylus tablet, a specific test from the large-scale public ChildCIdb database. A complete analysis of the proposed approach is carried out using different time series selection techniques to choose the most discriminative ones for the age group detection task: (i) a statistical analysis and (ii) an automatic algorithm called sequential forward search (SFS). In addition, different classification algorithms such as dynamic time warping barycenter averaging (DBA) and hidden Markov models (HMM) are studied. Accuracy results over 85% are achieved, outperforming previous approaches in the literature and in more challenging age group conditions. Finally, the approach presented in this study can benefit many children-related applications, for example, towards an age-appropriate environment with the technology.

Maintaining a healthy lifestyle has become increasingly challenging in today's sedentary society marked by poor eating habits. To address this issue, both national and international organisations have made numerous efforts to promote healthier diets and increased physical activity. However, implementing these recommendations in daily life can be difficult, as they are often generic and not tailored to individuals. This study presents the AI4Food-NutritionDB database, the first nutrition database that incorporates food images and a nutrition taxonomy based on recommendations by national and international health authorities. The database offers a multi-level categorisation, comprising 6 nutritional levels, 19 main categories (e.g., "Meat"), 73 subcategories (e.g., "White Meat"), and 893 specific food products (e.g., "Chicken"). The AI4Food-NutritionDB opens the doors to new food computing approaches in terms of food intake frequency, quality, and categorisation. Also, we present a standardised experimental protocol and benchmark including three tasks based on the nutrition taxonomy (i.e., category, subcategory, and final product recognition). These resources are available to the research community, including our deep learning models trained on AI4Food-NutritionDB, which can serve as pre-trained models, achieving accurate recognition results for challenging food image databases.

This work introduces an innovative method for estimating attention levels (cognitive load) using an ensemble of facial analysis techniques applied to webcam videos. Our method is particularly useful, among others, in e-learning applications, so we trained, evaluated, and compared our approach on the mEBAL2 database, a public multi-modal database acquired in an e-learning environment. mEBAL2 comprises data from 60 users who performed 8 different tasks. These tasks varied in difficulty, leading to changes in their cognitive loads. Our approach adapts state-of-the-art facial analysis technologies to quantify the users' cognitive load in the form of high or low attention. Several behavioral signals and physiological processes related to the cognitive load are used, such as eyeblink, heart rate, facial action units, and head pose, among others. Furthermore, we conduct a study to understand which individual features obtain better results, the most efficient combinations, explore local and global features, and how temporary time intervals affect attention level estimation, among other aspects. We find that global facial features are more appropriate for multimodal systems using score-level fusion, particularly as the temporal window increases. On the other hand, local features are more suitable for fusion through neural network training with score-level fusion approaches. Our method outperforms existing state-of-the-art accuracies using the public mEBAL2 benchmark.