arXiv:2511.14792v1 Announce Type: cross 
Abstract: Fiber Specklegram Sensors (FSS) are highly effective for environmental monitoring, particularly for detecting temperature variations. However, the nonlinear nature of specklegram data presents significant challenges for accurate temperature prediction. This study investigates the use of transformer-based architectures, including Vision Transformers (ViTs), Swin Transformers, and emerging models such as Learnable Importance Non-Symmetric Attention Vision Transformers (LINA-ViT) and Multi-Adaptive Proximity Vision Graph Attention Transformers (MAP-ViGAT), to predict temperature from specklegram data over a range of 0 to 120 Celsius. The results show that ViTs achieved a Mean Absolute Error (MAE) of 1.15, outperforming traditional models such as CNNs. GAT-ViT and MAP-ViGAT variants also demonstrated competitive accuracy, highlighting the importance of adaptive attention mechanisms and graph-based structures in capturing complex modal interactions and phase shifts in specklegram data. Additionally, this study incorporates Explainable AI (XAI) techniques, including attention maps and saliency maps, to provide insights into the decision-making processes of the transformer models, improving interpretability and transparency. These findings establish transformer architectures as strong benchmarks for optical fiber-based temperature sensing and offer promising directions for industrial monitoring and structural health assessment applications.

تستكشف هذه الدراسة تطبيق الهياكل المعتمدة على المحولات للتنبؤ بتغيرات درجة الحرارة باستخدام حساسات specklegram الألياف (FSS). تبرز الأبحاث التحديات التي تطرحها الطبيعة غير الخطية لبيانات specklegram وتظهر أن المحولات البصرية (ViTs) حققت متوسط خطأ مطلق (MAE) قدره 1.15، متفوقة على النماذج التقليدية مثل CNNs. تؤكد النتائج على إمكانيات النماذج المتقدمة من المحولات في تعزيز قدرات المراقبة البيئية.

Este estudio investiga la aplicación de arquitecturas basadas en transformadores para predecir variaciones de temperatura utilizando Sensores de Specklegrama de Fibra (FSS). La investigación destaca los desafíos que presenta la naturaleza no lineal de los datos de specklegrama y demuestra que los Transformadores de Visión (ViTs) lograron un Error Absoluto Medio (MAE) de 1.15, superando a modelos tradicionales como las CNN. Los hallazgos subrayan el potencial de los modelos avanzados de transformadores para mejorar las capacidades de monitoreo ambiental.

Cette étude examine l'application des architectures basées sur des transformateurs pour prédire les variations de température à l'aide de capteurs de specklegramme en fibre (FSS). La recherche met en évidence les défis posés par la nature non linéaire des données de specklegramme et démontre que les transformateurs de vision (ViTs) ont atteint une erreur absolue moyenne (MAE) de 1,15, surpassant les modèles traditionnels tels que les CNN. Les résultats soulignent le potentiel des modèles avancés de transformateurs pour améliorer les capacités de surveillance environnementale.

This study explores the application of transformer-based architectures for predicting temperature variations using Fiber Specklegram Sensors (FSS). The research highlights the challenges posed by the nonlinear nature of specklegram data and demonstrates that Vision Transformers (ViTs) achieved a Mean Absolute Error (MAE) of 1.15, outperforming traditional models like CNNs. The findings underscore the potential of advanced transformer models in enhancing environmental monitoring capabilities.

Application of Graph Based Vision Transformers Architectures for Accurate Temperature Prediction in Fiber Specklegram Sensors

arXiv:2601.08499v1 Announce Type: new 
Abstract: Large models such as Vision Transformers (ViTs) have demonstrated remarkable superiority over smaller architectures like ResNet in few-shot classification, owing to their powerful representational capacity. However, fine-tuning such large models demands extensive GPU memory and prolonged training time, making them impractical for many real-world low-resource scenarios. To bridge this gap, we propose EfficientFSL, a query-only fine-tuning framework tailored specifically for few-shot classification with ViT, which achieves competitive performance while significantly reducing computational overhead. EfficientFSL fully leverages the knowledge embedded in the pre-trained model and its strong comprehension ability, achieving high classification accuracy with an extremely small number of tunable parameters. Specifically, we introduce a lightweight trainable Forward Block to synthesize task-specific queries that extract informative features from the intermediate representations of the pre-trained model in a query-only manner. We further propose a Combine Block to fuse multi-layer outputs, enhancing the depth and robustness of feature representations. Finally, a Support-Query Attention Block mitigates distribution shift by adjusting prototypes to align with the query set distribution. With minimal trainable parameters, EfficientFSL achieves state-of-the-art performance on four in-domain few-shot datasets and six cross-domain datasets, demonstrating its effectiveness in real-world applications.

يقدم EfficientFSL إطارًا للتعديل يعتمد فقط على الاستفسارات لتحسين تصنيف القليل من العينات باستخدام المحولات البصرية (ViTs)، مما يعزز الأداء مع تقليل كبير في المتطلبات الحاسوبية. تستفيد هذه الطريقة من قدرات النموذج المدرب مسبقًا، محققة دقة عالية مع عدد قليل من المعلمات القابلة للتعديل.

EfficientFSL presenta un marco de ajuste solo de consultas para Transformadores de Visión (ViTs), mejorando la clasificación de pocos disparos mientras reduce significativamente las demandas computacionales. Este enfoque aprovecha las capacidades del modelo preentrenado, logrando alta precisión con un número mínimo de parámetros.

EfficientFSL introduit un cadre de réglage uniquement basé sur les requêtes pour les Transformateurs de Vision (ViTs), améliorant la classification à faible échantillon tout en réduisant considérablement les exigences computationnelles. Cette approche exploite les capacités du modèle pré-entraîné, atteignant une haute précision avec un nombre minimal de paramètres.

EfficientFSL introduces a query-only fine-tuning framework for Vision Transformers (ViTs), enhancing few-shot classification while significantly reducing computational demands. This approach leverages the pre-trained model's capabilities, achieving high accuracy with minimal parameters.

EfficientFSL: Enhancing Few-Shot Classification via Query-Only Tuning in Vision Transformers

arXiv:2601.08703v1 Announce Type: cross 
Abstract: Explainable artificial intelligence (XAI) is concerned with producing explanations indicating the inner workings of models. For a Rashomon set of similarly performing models, explanations provide a way of disambiguating the behavior of individual models, helping select models for deployment. However explanations themselves can vary depending on the explainer used, and need to be evaluated. In the paper "Evaluating Model Explanations without Ground Truth", we proposed three principles of explanation evaluation and a new method "AXE" to evaluate the quality of feature-importance explanations. We go on to illustrate how evaluation metrics that rely on comparing model explanations against ideal ground truth explanations obscure behavioral differences within a Rashomon set. Explanation evaluation aligned with our proposed principles would highlight these differences instead, helping select models from the Rashomon set. The selection of alternate models from the Rashomon set can maintain identical predictions but mislead explainers into generating false explanations, and mislead evaluation methods into considering the false explanations to be of high quality. AXE, our proposed explanation evaluation method, can detect this adversarial fairwashing of explanations with a 100% success rate. Unlike prior explanation evaluation strategies such as those based on model sensitivity or ground truth comparison, AXE can determine when protected attributes are used to make predictions.

يناقش المقال بعنوان 'تقييم قدرة الشروحات على توضيح النماذج في مجموعة راشومون' دور الذكاء الاصطناعي القابل للتفسير (XAI) في توضيح سلوك النماذج ضمن مجموعة راشومون، حيث تؤدي نماذج متعددة أداءً مشابهًا. يقدم طريقة AXE لتقييم شروحات أهمية الميزات، مؤكدًا على الحاجة إلى مقاييس تقييم فعالة تكشف عن الاختلافات السلوكية بين النماذج.

El artículo titulado 'Evaluando la capacidad de las explicaciones para desambiguar modelos en un conjunto de Rashomon' discute el papel de la inteligencia artificial explicable (XAI) en la clarificación del comportamiento de los modelos dentro de un conjunto de Rashomon, donde múltiples modelos tienen un rendimiento similar. Introduce el método AXE para evaluar explicaciones de importancia de características, enfatizando la necesidad de métricas de evaluación efectivas que revelen diferencias de comportamiento entre los modelos.

L'article intitulé 'Évaluation de la capacité des explications à disambiguïser les modèles dans un ensemble de Rashomon' aborde le rôle de l'intelligence artificielle explicable (XAI) dans la clarification du comportement des modèles au sein d'un ensemble de Rashomon, où plusieurs modèles ont des performances similaires. Il introduit la méthode AXE pour évaluer les explications d'importance des caractéristiques, en soulignant la nécessité de métriques d'évaluation efficaces qui révèlent les différences comportementales entre les modèles.

The paper titled 'Evaluating the Ability of Explanations to Disambiguate Models in a Rashomon Set' discusses the role of explainable artificial intelligence (XAI) in clarifying the behavior of models within a Rashomon set, where multiple models perform similarly. It introduces the AXE method for evaluating feature-importance explanations, emphasizing the need for effective evaluation metrics that reveal behavioral differences among models.

Application of Graph Based Vision Transformers Architectures for Accurate Temperature Prediction in Fiber Specklegram Sensors

Was this article worth reading? Share it

LucidQuery AI

The Visualizer

Attentive AI

Octofy

Graphite Note

VibeFrame

Ready to build your own newsroom?