This paper focuses on the many problems that exist in the evaluation of AI to improve the usability of Web applications for multilingual learning platforms, including the poor alignment effect between learning efficiency and usability, the imbalance of cognitive load regulation, and the singularity of evaluation indicators. In order to solve these problems, this study innovatively constructs a multi-dimensional evaluation index system integrating language cognitive load and learning efficiency and designs a dynamic evaluation model AILA-WA driven by AI. This model can combine learning algorithms to interact with data from Web applications and can collect real-time data related to language learning behavior and cognitive state feedback data from Web applications. It enables the optimization direction of Web applications to be identified and accurately quantified. Subsequent experiments successfully prove that the index system and the evaluation model can effectively improve the comprehensiveness accuracy of the evaluation of Web application usability. For example, in the scenario of multilingual learning, the cognitive load fitting deviation rate of the Web application using the AILA-model is the best compared to the Web application using the comparative model. At the same time, learning efficiency and CSAT user satisfaction are also at the level; and the model is suitable for Web applications. System response delay on multiple terminals is reduced to 0.3 s. These breakthroughs provide strong support for design iteration and usability optimization of AI to improve multilingual learning Web applications.

Distributed Acoustic Sensing (DAS) offers an innovative solution for intelligent transportation systems, leveraging its benefits of all-weather monitoring, extensive coverage, and low maintenance costs. However, most existing studies use dark fiber as sensing arrays, which results in critical issues such as insufficient resolution, poor time-domain continuity, and a low signal-to-noise ratio in the vehicle signals collected by DAS systems, thereby limiting the practical application of this technology in traffic monitoring. To address these technical bottlenecks, this study proposes a novel, rapidly deployable intelligent traffic monitoring system based on DAS. Firstly, the system utilizes the self-developed high spatial resolution interrogator, and by optimizing fiber cable selection and introducing the Variational Mode Decomposition (VMD) method, the adverse effects of the fiber stress relaxation on signal demodulation are effectively suppressed from both the fiber structure and digital signal processing perspectives. Secondly, based on this system, high-fidelity vehicle signals were successfully collected, achieving the first direct and accurate measurement of wheel count and wheelbase under the condition of optical fiber deployment parallel to the road. Lastly, the system also successfully captured characteristic signals of pedestrians in different motion states, with rich detailed information. This study reveals the significant potential of DAS technology in heterogeneous traffic features recognition and provides high-quality data support for the development of high-precision DAS traffic monitoring technology.

This study addresses the fine-grained segmentation of defects in bridge pavements, which is crucial for the maintenance and structural safety of bridges. Although bridge pavements pose distinctive challenges owing to their unique characteristics and varied defect types, previous studies have primarily focused on the detection of slender cracks. To fill this research gap, we developed a novel end-to-end hybrid method that dynamically combines the vision transformer (ViT) and level set theory to handle the complex geometry of bridge pavement defects. The novelty of the proposed method lies in the configuration of two parallel decoders. These decoders, operating under a unified objective function, share weights and perform simultaneous optimization, thereby facilitating a holistic end-to-end training process. Furthermore, we compiled two new bridge pavement defect datasets, namely BdridgeDefX and BdridgeDef20, which offer broader applicability for practical defect detection. The results of a rigorous experimental validation on four datasets demonstrated the proposed method’s capability of generating accurate defect boundaries and delivering state-of-the-art performance.

In the field of the Internet of Things, sensors with large-scale perception ability are urgently needed for the scenarios, such as structure health monitoring, vehicle tracking, and so on. Quasi-distributed acoustic sensing (QDAS), which uses a sensing fiber with an ultraweak reflector array to perceive thousands of points at the same time, shows great advantages in these applications. The sensing fiber integrates sensing and transmission functionalities together and is easy to be deployed on large-scale targets. With the development of QDAS, the performance limitations caused by the finite frequency-domain resource are gradually emerging. Multiplexing the frequency-domain resource can be the key to further performance breakthroughs. For this purpose, the multiple-input–multiple-output (MIMO) technology, which has been widely used in communication/radar systems, is introduced into QDAS in this article. The principles of QDAS with the MIMO technology are elaborated, and the feasibility is verified through simulations and experiments, in which the response bandwidth is tripled compared to the traditional single-pulse QDAS. To the best of our knowledge, this is the first time that the MIMO coding technology, i.e., orthogonal codes with the same frequency (OCSF), is used in QDAS. This work paves a new way for breaking QDAS performance limitations that are bounded by the finite frequency resource.

Phase-sensitive optical time-domain reflectometry (Φ-OTDR), which utilizes the phase information of Rayleigh scattered lightwave inside optical fiber, could turn a fiber cable into a massive sensor array for distributed acoustic sensing (DAS), i.e., an emerging infrastructure for Internet of Things. Given a certain fiber length, there are tradeoffs among the sensing bandwidth, the sensitivity, and the spatial resolution. In this paper, the concept of linearization and Golay pulse-coding for heterodyne Φ-OTDR are proposed and experimentally verified for the first time. First, we gave a full theoretical treatment on how an intensity-coded yet phase-retrieved Φ-OTDR can be built up as a fully linear system, therefore a significant enhancement of signal-to-noise ratio becomes viable and the sensing bandwidth equals the four-times averaging case. Second in the proof-of-concept experiment, submeter gauge length and nanostrain resolution were realized with 10 km sensing range, in other words, more than ten thousand sensitive sensing units were realized along the fiber. This paper makes a significant step toward high-performance DAS with orders-of-magnitude performance enhancement.