AI, Volume 6, Issue 4 (April 2025) – 24 articles

A central focus of this study was the methodology used to evaluate both humans and AI platforms, particularly in terms of their competitiveness and the implications of six key challenges to society resulting from the development and increasing use of artificial intelligence (AI) technologies. The list of challenges was compiled by consulting various online sources and cross-referencing with academics from 15 countries across Europe and the USA. Professors, scientific researchers, and PhD students were invited to independently and remotely evaluate the challenges. Rather than contributing another discussion based solely on social arguments, this paper seeks to provide a logical evaluation framework, moving beyond qualitative discourse by incorporating numerical values. The pairwise comparison of AI challenges was conducted by two groups of participants using the multicriteria decision-making model known as the analytic hierarchy process (AHP). Thirty-eight humans performed pairwise comparisons of the six challenges after they were listed in a distributed questionnaire. The same procedure was carried out by four AI platforms—ChatGPT, Gemini (BardAI), Perplexity, and DedaAI—who responded to the same requests as the human participants. The results from both groups were grouped and compared, revealing interesting differences in the prioritization of AI challenges’ impact on society. Both groups agreed on the highest importance of data privacy and security, as well as the lowest importance of social and cultural resistance, specifically the clash of AI with existing cultural norms and societal values. Full article

Efficiently handling long contexts in transformer-based language models with low perplexity is an active area of research. Numerous recent approaches like Linformer, Longformer, Performer, and Structured state space models (SSMs), have not fully resolved this problem. All these models strive to reduce the quadratic time complexity of the attention mechanism while minimizing the loss in quality due to the effective compression of the long context. Inspired by the cache and virtual memory principle in computers, where in case of a cache miss, not only the needed data are retrieved from the memory, but the adjacent data are also obtained, we apply this concept to handling long contexts by dividing it into small segments. In our design, we retrieve the nearby segments in an uncompressed form when high segment-level attention occurs at the compressed level. Our enhancements for handling long context include aggregating four attention mechanisms consisting of short sliding window attention, long compressed segmented attention, dynamically retrieving top-k high-attention uncompressed segments, and overlapping segments in long segment attention to avoid segment fragmentation. These enhancements result in an architecture that outperforms existing SOTA architectures with an average perplexity improvement of 8.5% over similar model sizes. Full article

Background/Objectives: Artificial intelligence (AI) is increasingly influencing oncological research by enabling precision medicine in ovarian cancer through enhanced prediction of therapy response and patient stratification. This systematic review and meta-analysis was conducted to assess the performance of AI-driven models across three key domains: genomics and molecular profiling, radiomics-based imaging analysis, and prediction of immunotherapy response. Methods: Relevant studies were identified through a systematic search across multiple databases (2020–2025), adhering to PRISMA guidelines. Results: Thirteen studies met the inclusion criteria, involving over 10,000 ovarian cancer patients and encompassing diverse AI models such as machine learning classifiers and deep learning architectures. Pooled AUCs indicated strong predictive performance for genomics-based (0.78), radiomics-based (0.88), and immunotherapy-based (0.77) models. Notably, radiogenomics-based AI integrating imaging and molecular data yielded the highest accuracy (AUC = 0.975), highlighting the potential of multi-modal approaches. Heterogeneity and risk of bias were assessed, and evidence certainty was graded. Conclusions: Overall, AI demonstrated promise in predicting therapeutic outcomes in ovarian cancer, with radiomics and integrated radiogenomics emerging as leading strategies. Future efforts should prioritize explainability, prospective multi-center validation, and integration of immune and spatial transcriptomic data to support clinical implementation and individualized treatment strategies. Unlike earlier reviews, this study synthesizes a broader range of AI applications in ovarian cancer and provides pooled performance metrics across diverse models. It examines the methodological soundness of the selected studies and highlights current gaps and opportunities for clinical translation, offering a comprehensive and forward-looking perspective in the field. Full article

Mind wandering is a common issue among schoolchildren and academic students, often undermining the quality of learning and teaching effectiveness. Current detection methods mainly rely on eye trackers and electrodermal activity (EDA) sensors, focusing on external indicators such as facial movements but neglecting voice detection. These methods are often cumbersome, uncomfortable for participants, and invasive, requiring specialized, expensive equipment that disrupts the natural learning environment. To overcome these challenges, a new algorithm has been developed to detect mind wandering during reading aloud. Based on external indicators like the blink rate, pitch frequency, and reading rate, the algorithm integrates these three criteria to ensure the accurate detection of mind wandering using only a standard computer camera and microphone, making it easy to implement and widely accessible. An experiment with ten participants validated this approach. Participants read aloud a text of 1304 words while the algorithm, incorporating the Viola–Jones model for face and eye detection and pitch-frequency analysis, monitored for signs of mind wandering. A voice activity detection (VAD) technique was also used to recognize human speech. The algorithm achieved 76% accuracy in predicting mind wandering during specific text segments, demonstrating the feasibility of using noninvasive physiological indicators. This method offers a practical, non-intrusive solution for detecting mind wandering through video and audio data, making it suitable for educational settings. Its ability to integrate seamlessly into classrooms holds promise for enhancing student concentration, improving the teacher–student dynamic, and boosting overall teaching effectiveness. By leveraging standard, accessible technology, this approach could pave the way for more personalized, technology-enhanced education systems. Full article

Three-dimensional (3D) visual perception is pivotal for understanding surrounding environments in applications such as autonomous driving and mobile robotics. While LiDAR-based models dominate due to accurate depth sensing, their cost and sparse outputs have driven interest in camera-based systems. However, challenges like cross-domain degradation and depth estimation inaccuracies persist. This paper introduces BEVCAM3D, a unified bird’s-eye view (BEV) architecture that fuses monocular cameras and LiDAR point clouds to overcome single-sensor limitations. BEVCAM3D integrates a deformable cross-modality attention module for feature alignment and a fast ground segmentation algorithm to reduce computational overhead by 40%. Evaluated on the nuScenes dataset, BEVCAM3D achieves state-of-the-art performance, with a 73.9% mAP and a 76.2% NDS, outperforming existing LiDAR-camera fusion methods like SparseFusion (72.0% mAP) and IS-Fusion (73.0% mAP). Notably, it excels in detecting pedestrians (91.0% AP) and traffic cones (89.9% AP), addressing the class imbalance in autonomous driving scenarios. The framework supports real-time inference at 11.2 FPS with an EfficientDet-B3 backbone and demonstrates robustness under low-light conditions (62.3% nighttime mAP). Full article

Background/Objectives: This article explores the use of Ancient Greek as a prompt language in DALL·E 3, an Artificial Intelligence software for image generation. The research investigates three dimensions of Artificial Intelligence’s ability: (a) the sense and visualization of the concept of distance, (b) the mixing of representational as well as mythic contents, and (c) the visualization of emotions. More specifically, the research not only investigates AI’s potentialities in processing and representing Ancient Greek texts but also attempts to assess its interpretative boundaries. The key question is whether AI can faithfully represent the underlying conceptual and narrative structures of ancient literature or whether its representations are superficial and constrained by algorithmic procedures. Methods: This is a mixed-methods experimental research design examining whether a specified Artificial Intelligence software can sense, understand, and graphically represent linguistic and conceptual structures in the Ancient Greek language. Results: The study highlights Artificial Intelligence’s possibility in classical language education as well as digital humanities regarding linguistic complexity versus AI’s power in interpretation. More specifically, the research not only investigates AI’s potentialities in processing and representing Ancient Greek texts but also attempts to assess its interpretative boundaries. The key question is whether AI can faithfully represent the underlying conceptual and narrative structures of ancient literature or whether its representations are superficial and constrained by algorithmic procedures. The study highlights Artificial Intelligence’s possibility in classical language education as well as digital humanities regarding linguistic complexity versus AI’s power in interpretation. Conclusions: The research is a step toward a more extensive discussion on Artificial Intelligence in historical linguistics, digital pedagogy, as well as aesthetic representation by Artificial Intelligence environments. Full article

(1) Acute pancreatitis (AP) is a medical emergency associated with high mortality rates. Early and accurate prognosis assessment during admission is crucial for optimizing patient management and outcomes. This study seeks to develop robust radiomics-based machine learning (ML) models to classify the severity of AP using contrast-enhanced computed tomography (CECT) scans. (2) Methods: A retrospective cohort of 287 AP patients with CECT scans was analyzed, and clinical data were collected within 72 h of admission. Patients were classified as mild or moderate/severe based on the Revised Atlanta classification. Two radiologists manually segmented the pancreas and peripancreatic regions on CECT scans, and 234 radiomic features were extracted. The performance of the ML algorithms was compared with that of traditional scoring systems, including Ranson and Glasgow-Imrie scores. (3) Results: Traditional severity scoring systems produced AUC values of 0.593 (Ranson, Admission), 0.696 (Ranson, 48 h), 0.677 (Ranson, Cumulative), and 0.663 (Glasgow-Imrie). Using LASSO regression, 12 radiomic features were selected for the ML classifiers. Among these, the best-performing ML classifier achieved an AUC of 0.876 in the training set and 0.777 in the test set. (4) Conclusions: Radiomics-based ML classifiers significantly enhanced the prediction of AP severity in patients undergoing CECT scans within 72 h of admission, outperforming traditional severity scoring systems. This research is the first to successfully predict prognosis by analyzing radiomic features from both pancreatic and peripancreatic tissues using multiple ML algorithms applied to early CECT images. Full article

Background/Objectives: This article addresses the challenge of stress detection across diverse contexts. Mental stress is a worldwide concern that substantially affects human health and productivity, rendering it a critical research challenge. Although numerous studies have investigated stress detection through machine learning (ML) techniques, there has been limited research on assessing ML models trained in one context and utilized in another. The objective of ML-based stress detection systems is to create models that generalize across various contexts. Methods: This study examines the generalizability of ML models employing EEG recordings from two stress-inducing contexts: mental arithmetic evaluation (MAE) and virtual reality (VR) gaming. We present a data collection workflow and publicly release a portion of the dataset. Furthermore, we evaluate classical ML models and their generalizability, offering insights into the influence of training data on model performance, data efficiency, and related expenses. EEG data were acquired leveraging MUSE-S^TM hardware during stressful MAE and VR gaming scenarios. The methodology entailed preprocessing EEG signals using wavelet denoising mother wavelets, assessing individual and aggregated sensor data, and employing three ML models—linear discriminant analysis (LDA), support vector machine (SVM), and K-nearest neighbors (KNN)—for classification purposes. Results: In Scenario 1, where MAE was employed for training and VR for testing, the TP10 electrode attained an average accuracy of 91.42% across all classifiers and participants, whereas the SVM classifier achieved the highest average accuracy of 95.76% across all participants. In Scenario 2, adopting VR data as the training data and MAE data as the testing data, the maximum average accuracy achieved was 88.05% with the combination of TP10, AF8, and TP9 electrodes across all classifiers and participants, whereas the LDA model attained the peak average accuracy of 90.27% among all participants. The optimal performance was achieved with Symlets 4 and Daubechies-2 for Scenarios 1 and 2, respectively. Conclusions: The results demonstrate that although ML models exhibit generalization capabilities across stressors, their performance is significantly influenced by the alignment between training and testing contexts, as evidenced by systematic cross-context evaluations using an 80/20 train–test split per participant and quantitative metrics (accuracy, precision, recall, and F1-score) averaged across participants. The observed variations in performance across stress scenarios, classifiers, and EEG sensors provide empirical support for this claim. Full article

Detecting fake news has become a critical challenge in today’s information-dense society. Existing research on fake news detection predominantly emphasizes multi-modal approaches, focusing primarily on textual and visual features. However, despite its clear importance, the integration of social context has received limited attention in the literature. To address this gap, this study proposes a novel three-dimensional multimodal fusion framework that integrates textual, visual, and social context features for effective fake news detection on social media platforms. The proposed methodology leverages an advanced Vision-and-Language Bidirectional Encoder Representations from Transformers multi-task model to extract fused attention features from text and images concurrently, capturing intricate inter-modal correlations. Comprehensive experiments validate the efficacy of the proposed approach. The results demonstrate that the proposed solution achieves the highest balanced accuracy of 77%, surpassing other baseline models. Furthermore, the incorporation of social context features significantly enhances model performance. The proposed multimodal architecture also outperforms state-of-the-art approaches, providing a robust and scalable framework for fake news detection using artificial intelligence. This study contributes to advancing the field by offering a comprehensive and practical engineering solution for combating fake news. Full article

The growing demand for telemedicine has highlighted the need for automated healthcare services, particularly in medical question classification. This study presents a deep learning model designed to address key challenges in telemedicine, including class imbalance and accurate routing of Arabic medical questions to the correct specialties. The model combines AraBERTv0.2-Twitter, fine-tuned for informal Arabic, with Bidirectional Long Short-Term Memory (BiLSTM) networks to capture deep semantic relationships in medical text. We used a labeled dataset of 5000 Arabic consultation records from Altibbi, covering five key medical specialties selected for their clinical relevance and frequency. The data underwent preprocessing to remove noise and normalize text. We employed stratified sampling to ensure representative distribution across the selected medical specialties. We evaluate multiple models using macro precision, macro recall, macro F1-score, weighted F1-score, and G-Mean. Our results demonstrate that DeepSMOTE combined with cross-entropy loss achieves the best performance. The findings offer statistically significant improvements and have practical implications for improving screening and patient routing in telemedicine platforms. Full article

Breast cancer detection is a critical task in healthcare, requiring fast, accurate, and efficient diagnostic tools. However, the high computational demands and latency of deep learning models in medical imaging present significant challenges, especially in resource-constrained environments. This paper addresses these challenges by presenting an FPGA hardware accelerator tailored for breast cancer classification, leveraging the Zynq XC7Z020 SoC. The system integrates FPGA-accelerated layers with an ARM Cortex-A9 processor to optimize both performance and resource efficiency. We developed modular IP cores, including Conv2D, Average Pooling, and ReLU, using Vivado HLS to maximize FPGA resource utilization. By adopting 8-bit fixed-point arithmetic, the design achieves a 15.8% reduction in execution time compared to traditional CPU-based implementations while maintaining high classification accuracy. Additionally, our optimized approach significantly enhances energy efficiency, reducing power consumption from 3.8 W to 1.4 W a 63.15% reduction. This improvement makes our design highly suitable for real-time, power-sensitive applications, particularly in embedded and edge computing environments. Furthermore, it underscores the scalability and efficiency of FPGA-based AI solutions for healthcare diagnostics, enabling faster and more energy-efficient deep learning inference on resource-constrained devices. Full article

The rise of large-scale language models and multimodal transformers has enabled instruction-based policies, such as vision-and-language navigation. To leverage their general world knowledge, we propose multimodal annotations for action options and support selection from a dynamic, describable action space. Our framework employs a multimodal transformer that processes front-facing camera images, light detection and ranging (LIDAR) sensor’s point clouds, and tasks as textual instructions to produce a history-aware decision policy for mobile robot navigation. Our approach leverages a pretrained vision–language encoder and integrates it with a custom causal generative pretrained transformer (GPT) decoder to predict action sequences within a state–action history. We propose a trainable attention score mechanism to efficiently select the most suitable action from a variable set of possible options. Action options are text–image pairs and encoded using the same multimodal encoder employed for environment states. This approach of annotating and dynamically selecting actions is applicable to broader multidomain decision-making tasks. We compared two baseline models, ViLT (vision-and-language transformer) and FLAVA (foundational language and vision alignment), and found that FLAVA achieves superior performance within the constraints of 8 GB video memory usage in the training phase. Experiments were conducted in both simulated and real-world environments using our custom datasets for instructed task completion episodes, demonstrating strong prediction accuracy. These results highlight the potential of multimodal, dynamic action spaces for instruction-based robot navigation and beyond. Full article

Beautimeter is a new tool powered by generative pre-trained transformer (GPT) technology, designed to evaluate architectural and urban beauty. Rooted in Christopher Alexander’s theory of centers, this work builds on the idea that all environments possess, to varying degrees, an innate sense of life. Alexander identified 15 fundamental properties, such as levels of scale and thick boundaries, that characterize living structure, which Beautimeter uses as a basis for its analysis. By integrating GPT’s advanced natural language processing capabilities, Beautimeter assesses the extent to which a structure embodies these 15 properties, enabling a nuanced evaluation of architectural and urban aesthetics. Using ChatGPT4o, the tool helps users generate insights into the perceived beauty and coherence of spaces. We conducted a series of case studies, evaluating images of architectural and urban environments, as well as carpets, paintings, and other artifacts. The results demonstrate Beautimeter’s effectiveness in analyzing aesthetic qualities across diverse contexts. Our findings suggest that by leveraging GPT technology, Beautimeter offers architects, urban planners, and designers a powerful tool to create spaces that resonate deeply with people. This paper also explores the implications of such technology for architecture and urban design, highlighting its potential to enhance both the design process and the assessment of built environments. Full article

The banking industry faces significant challenges, from high customer churn rates to threatening long-term revenue generation. Traditionally, churn models assess service quality using customer satisfaction metrics; however, these subjective variables often yield low predictive accuracy. This study examines the relationship between customer attrition and account balance using decision trees (DT), random forests (RF), and gradient-boosting machines (GBM). This research utilises a customer churn dataset and applies synthetic oversampling to balance class distribution during the preprocessing of financial variables. Account balance service is the primary factor in predicting customer churn, as it yields more accurate predictions compared to traditional subjective assessment methods. The tested model set achieved its highest predictive performance by applying boosting methods. The evaluation of research data highlights the critical role of financial indicators in shaping effective customer retention strategies. By leveraging machine learning intelligence, banks can make more informed decisions, attract new clients, and mitigate churn risk, ultimately enhancing long-term financial results. Full article

In traffic safety analysis, previous research has often focused on tabular data or textual crash narratives in isolation, neglecting the potential benefits of a hybrid multimodal approach. This study introduces the Multimodal Data Fusion (MDF) framework, which fuses tabular data with textual narratives by leveraging advanced Large Language Models (LLMs), such as GPT-2, GPT-3.5, and GPT-4.5, using zero-shot (ZS), few-shot (FS), and fine-tuning (FT) learning strategies. We employed few-shot learning with GPT-4.5 to generate new labels for traffic crash analysis, such as driver fault, driver actions, and crash factors, alongside the existing label for severity. Our methodology was tested on crash data from the Missouri State Highway Patrol, demonstrating significant improvements in model performance. GPT-2 (fine-tuned) was used as the baseline model, against which more advanced models were evaluated. GPT-4.5 few-shot learning achieved 98.9% accuracy for crash severity prediction and 98.1% accuracy for driver fault classification. In crash factor extraction, GPT-4.5 few-shot achieved the highest Jaccard score (82.9%), surpassing GPT-3.5 and fine-tuned GPT-2 models. Similarly, in driver actions extraction, GPT-4.5 few-shot attained a Jaccard score of 73.1%, while fine-tuned GPT-2 closely followed with 72.2%, demonstrating that task-specific fine-tuning can achieve performance close to state-of-the-art models when adapted to domain-specific data. These findings highlight the superior performance of GPT-4.5 few-shot learning, particularly in classification and information extraction tasks, while also underscoring the effectiveness of fine-tuning on domain-specific datasets to bridge performance gaps with more advanced models. The MDF framework’s success demonstrates its potential for broader applications beyond traffic crash analysis, particularly in domains where labeled data are scarce and predictive modeling is essential. Full article

Marine mammal monitoring, a growing field of research, is critical to cetacean conservation. Traditional ‘tagging’ attaches sensors such as GPS to such animals, though these are intrusive and susceptible to infection and, ultimately, death. A less intrusive approach exploits UUV commanded by a human operator above ground. The development of AI for autonomous underwater vehicle navigation models training environments in simulation, providing visual and physical fidelity suitable for sim-to-real transfer. Previous solutions, including UVMS and L2D, provide only satisfactory results, due to poor environment generalisation while sensors including sonar create environmental disturbances. Though rich in features, image data suffer from high dimensionality, providing a state space too great for many machine learning tasks. Underwater environments, susceptible to image noise, further complicate this issue. We propose SWiMM_2.0, coupling a Unity simulation modelling of a BLUEROV UUV with a DRL backend. A pre-processing step exploits a state-of-the-art CMVAE, reducing dimensionality while minimising data loss. Sim-to-real generalisation is validated by prior research. Custom behaviour metrics, unbiased to the naked eye and unprecedented in current ROV simulators, link our objectives ensuring successful ROV behaviour while tracking targets. Our experiments show that SAC maximises the former, achieving near-perfect behaviour while exploiting image data alone. Full article

Efficient pedestrian detection plays an important role in many practical daily life applications, such as autonomous cars, video surveillance, and intelligent driving assistance systems. The main goal of pedestrian detection systems, especially in vehicles, is to prevent accidents. By recognizing pedestrians in real time, these systems can alert drivers or even autonomously apply brakes, minimizing the possibility of collisions. However, occlusion is a major obstacle to pedestrian detection. Pedestrians are typically occluded by trees, street poles, cars, and other pedestrians. State-of-the-art detection methods are based on fully visible or little-occluded pedestrians; hence, their performance declines with increasing occlusion level. To meet this challenge, a pedestrian detector capable of handling occlusion is preferred. To increase the detection accuracy for occluded pedestrians, we propose a new method called the Discriminative Deformable Part Model (DDPM), which uses the concept of breaking human image into deformable parts via machine learning. In existing works, human image breaking into deformable parts has been performed by human intuition. In our novel approach, machine learning is used for deformable objects such as humans, combining the benefits and removing the drawbacks of the previous works. We also propose a new pedestrian dataset based on Eastern clothes to accommodate the detector’s evaluation under different intra-class variations of pedestrians. The proposed method achieves a higher detection accuracy on Pascal VOC and VisDrone Detection datasets when compared with other popular detection methods. Full article

Anti-money laundering (AML) systems are important for protecting the global economy. However, conventional rule-based methods rely on domain knowledge, leading to suboptimal accuracy and a lack of scalability. Graph neural networks (GNNs) for digraphs (directed graphs) can be applied to transaction graphs and capture suspicious transactions or accounts. However, most spectral GNNs do not naturally support multi-dimensional edge features, lack interpretability due to edge modifications, and have limited scalability owing to their spectral nature. Conversely, most spatial methods may not capture the money flow well. Therefore, in this work, we propose LineMVGNN (Line-Graph-Assisted Multi-View Graph Neural Network), a novel spatial method that considers payment and receipt transactions. Specifically, the LineMVGNN model extends a lightweight MVGNN module, which performs two-way message passing between nodes in a transaction graph. Additionally, LineMVGNN incorporates a line graph view of the original transaction graph to enhance the propagation of transaction information. We conduct experiments on two real-world account-based transaction datasets: the Ethereum phishing transaction network dataset and a financial payment transaction dataset from one of our industry partners. The results show that our proposed method outperforms state-of-the-art methods, reflecting the effectiveness of money laundering detection with line-graph-assisted multi-view graph learning. We also discuss scalability, adversarial robustness, and regulatory considerations of our proposed method. Full article

Voice data contain a wealth of temporal and spectral information and can be a valuable resource for disease classification. However, traditional methods are often not effective in capturing the key features required for the classification of multiple disease classes. To address this challenge, we propose a voice-based multi-disease detection approach with a lightweight attention-based temporal convolution neural network (Voice-AttentionNet) designed to analyze speech data for multi-class disease classification. Our model utilizes the temporal convolution neural network (CNN) architecture to extract high-resolution temporal features, while incorporating attention mechanisms to highlight disease-related patterns. Extensive experiments have been conducted on our dataset, including speech samples from patients with multiple illnesses. The results show that our method achieves the most advanced performance with an average classification accuracy of 91.61% on six datasets and is superior to the existing classical models. These findings highlight the potential of combining attention mechanisms with temporal CNNs in the use of speech data for disease classification. Moreover, this study provides a promising direction for deploying AI-driven diagnostic tools in clinical scenarios. Full article

Feature retrieval technology for building floor plans has garnered significant attention in recent years due to its critical role in the efficient management and execution of construction projects. This paper presents a comprehensive exploration of four primary features essential for the retrieval of building floor plans: semantic features, spatial features, shape features, and texture features (collectively referred to as 3ST features). The extraction algorithms and underlying principles associated with these features are thoroughly analyzed, with a focus on advanced methods such as wavelet transforms and Fourier shape descriptors. Furthermore, the performance of various retrieval algorithms is evaluated through rigorous experimental analysis, offering valuable insights into optimizing the retrieval of building floor plans. Finally, this study outlines prospective directions for the advancement of feature retrieval technology in the context of floor plans. Full article

Large Language Models (LLMs) have immensely advanced the field of Artificial Intelligence (AI), with recent models being able to perform chain-of-thought reasoning and solve complex mathematical problems, ranging from theorem proving to ones involving advanced calculus. The success of LLMs derives from a combination of the Transformer architecture with its attention mechanism, the autoregressive training methodology with masked attention, and the alignment fine-tuning via reinforcement learning algorithms. In this research, we attempt to explore a possible solution to the fundamental NP-hard problem of combinatorial optimization, in particular, the Traveling Salesman Problem (TSP), by following the LLM approach in terms of the architecture and training algorithms. Similar to the LLM design, which is trained in an autoregressive manner to predict the next token, our model is trained to predict the next node in a TSP graph. After the model is trained on random TSP graphs with known near-optimal solutions, we fine-tune the model using Direct Preference Optimization (DPO). The tour generation in a trained model is autoregressive one-step generation with no need for iterative refinement. Our results are very promising and indicate that, for TSP graphs up to 100 nodes, a relatively small amount of training data yield solutions within a few percent of the optimal. This optimization improves if more data are used to train the model. Full article

Natural Language Processing (NLP) and advanced acoustic analysis have opened new avenues in animal welfare research by decoding the vocal signals of farm animals. This study explored the feasibility of adapting a large-scale Transformer-based model, OpenAI’s Whisper, originally developed for human speech recognition, to decode chicken vocalizations. Our primary objective was to determine whether Whisper could effectively identify acoustic patterns associated with emotional and physiological states in poultry, thereby enabling real-time, non-invasive welfare assessments. To achieve this, chicken vocal data were recorded under diverse experimental conditions, including healthy versus unhealthy birds, pre-stress versus post-stress scenarios, and quiet versus noisy environments. The audio recordings were processed through Whisper, producing text-like outputs. Although these outputs did not represent literal translations of chicken vocalizations into human language, they exhibited consistent patterns in token sequences and sentiment indicators strongly correlated with recognized poultry stressors and welfare conditions. Sentiment analysis using standard NLP tools (e.g., polarity scoring) identified notable shifts in “negative” and “positive” scores that corresponded closely with documented changes in vocal intensity associated with stress events and altered physiological states. Despite the inherent domain mismatch—given Whisper’s original training on human speech—the findings clearly demonstrate the model’s capability to reliably capture acoustic features significant to poultry welfare. Recognizing the limitations associated with applying English-oriented sentiment tools, this study proposes future multimodal validation frameworks incorporating physiological sensors and behavioral observations to further strengthen biological interpretability. To our knowledge, this work provides the first demonstration that Transformer-based architectures, even without species-specific fine-tuning, can effectively encode meaningful acoustic patterns from animal vocalizations, highlighting their transformative potential for advancing productivity, sustainability, and welfare practices in precision poultry farming. Full article

With the rise of e-commerce and web application usage, recommendation systems have become important to our daily tasks. They provide personalized suggestions to assist with any task under consideration. While various machine learning algorithms have been developed for recommendation tasks, existing systems still face limitations. This research focuses on advancing context-aware recommendation sytems by leveraging the capabilities of Large Language Models (LLMs) in conjunction with real-time data. The research exploits the integration of existing real-time data APIs with LLMs to enhance the capabilities of the recommendation systems already integrated into smart societies. The experimental results demonstrate that the hybrid approach significantly improves the user experience and recommendation quality, ensuring more relevant and dynamic suggestions. Full article

Birds can cause substantial damage to crops, directly affecting farmers’ productivity and profitability. As a result, detecting bird presence in crop fields is crucial for effective crop management. Traditional agricultural practices have used various tools and techniques to deter pest birds, while digital agriculture has advanced these efforts through Internet of Things (IoT) and artificial intelligence (AI) technologies. With recent advancements in hardware and processing chips, connected devices can now utilize deep convolutional neural networks (CNNs) for on-field image classification. However, training these models can be energy-intensive, especially when large amounts of data, such as images, need to be transmitted for centralized model training. Federated learning (FL) offers a solution by enabling local training on edge devices, reducing data transmission costs and energy demands while also preserving data privacy and achieving shared model knowledge across connected devices. This paper proposes a low-energy federated learning framework for a compact smart camera network designed to perform simple image classification for bird detection in crop fields. The results demonstrate that this decentralized approach achieves performance comparable to a centrally trained model while consuming at least 8 times less energy. Further efficiency improvements, with a minimal tradeoff in performance reduction, are explored through early stopping. Full article