Abstract
Despite their extreme sensitivity, speckle-based fiber optical sensors are typically limited by the camera frame rate and dynamic range. In this context, recent developments in event-based sensors make them a promising and affordable tool for high-speed interrogation for such class of sensors, offering a low-latency approach to detecting dynamic changes in illumination patterns, well-suited for fast interrogation with frequency response up to the MHz range. In this manuscript, we investigate the potential of using an event-based vision sensor (EVS) as an interrogator for a speckle-based optical fiber sensor operating at 532nm to detect vibrations induced by an off-the-shelf sound speaker. In contact with the fiber, these vibrations induce dynamic changes in the speckle pattern, which are tracked by the EVS and processed to construct temporal frames with timestamps below 100 mu s. Approximating the differential operator of the deformation in the linear regime, we show a successful reconstruction of the acoustic signal for two illustrative case studies: i)a single-frequency signal at 1.2 KHz and ii)a linear ramp between 300 Hz to 2.5 kHz. The results demonstrate the ability to accurately identify not only the fundamental frequencies but also their harmonics generated by the speaker up to 5 KHz, paving an innovative path to harness the potential of speckle-based sensors in multiple scenarios of optical metrology and dynamic sensing applications.

Abstract
Software testing is an integral part of modern software development. Testing frameworks are part of the toolset of any software language allowing programmers to test their programs in order to detect bugs. Unfortunately, there is no work on testing in attribute grammars. In this paper we combine the powerful property-based testing technique with the attribute grammar formalism. In such property-based attribute grammars, properties are defined on attribute instances. Properties are tested on large sets of randomly generated (abstract syntax) trees by evaluating their attributes. We present an implementation that relies on strategies to express property-based attribute grammars. Strategies are tree-based recursion patterns that are used to encode logic quantifiers defining the properties.

Abstract
Underwater environments pose unique challenges to optical systems due to physical phenomena that induce severe data degradation. Current imaging sensors rarely address these effects comprehensively, resulting in the need to integrate complementary information sources. This article presents a multimodal data fusion approach to combine information from diverse sensing modalities into a single dense and accurate tridimensional representation. The proposed fusiNg tExture with apparent motion information for underwater Scene recOnstruction (NESO) encoder-decoder network leverages motion perception principles to extract relative depth cues, fusing them with textured information through an early fusion strategy. Evaluated on the FLSea-Stereo dataset, NESO outperforms state-of-the-art methods by 58.7%. Dense depth maps are achieved using multi-stage skip connections with attention mechanisms that ensure propagation of key features across network levels. This representation is further enhanced by incorporating sparse but millimeter-precise depth measurements from active imaging techniques. A regression-based algorithm maps depth displacements between these heterogeneous point clouds, using the estimated curves to refine the dense NESO prediction. This approach achieves relative errors as low as 0.41% when reconstructing submerged anode structures, accounting for metric improvements of up to 0.1124 m relative to the initial measurements. Validation at the ATLANTIS Coastal Testbed demonstrates the effectiveness of this multimodal fusion approach in obtaining robust tri-dimensional representations in real underwater conditions.

Abstract
When directional data fall in the positive orthant of the unit hypersphere, a folded directional distribution is preferred over a simple directional distribution for modeling the data. Since directional data, especially axial data, can be modeled using a Watson distribution, this paper considers a folded Watson distribution for such cases. We first address the parameter estimation of this distribution using maximum likelihood, which requires a numerical algorithm to solve the likelihood equations. We use the Expectation-Maximization (EM) algorithm to obtain these estimates and to analyze the properties of the concentration estimator through simulation. Next, we propose the Bayes rule for a folded Watson distribution and evaluate its performance through simulation in various scenarios, comparing it with the Bayes rule for the Watson distribution. Finally, we present examples using both simulated and real data available in the literature.

Abstract
Personal data privacy is fundamental in human activity and health monitoring systems, with additional challenges posed by the integration of AI tools. For monitoring to be effective, the user needs to trust on the system, adopt and use it frequently. Besides data privacy requirements and regulatory compliance, transparency, explainability and accountability matter. By incorporating Privacy by Design principles into AI-driven systems to ensure GDPR alignment, this paper proposed a simple approach for embedding privacy-preserving mechanisms throughout the data lifecycle, from design to deployment and continuous monitoring and illustrate it with two use cases developing AI-Driven Systems for human activity and health monitoring in different contexts. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

Abstract