98047 research outputs found
Sort by
Development and Deployment of a Virtual Water Gauge System Utilizing the ResNet-50 Convolutional Neural Network for Real-Time River Water Level Monitoring: A Case Study of the Keelung River in Taiwan
No full textClimate change has exacerbated severe rainfall events, leading to rapid and unpredictable fluctuations in river water levels. This environment necessitates the development of real-time, automated systems for water level detection. Due to degradation, traditional methods relying on physical river gauges are becoming progressively unreliable. This paper presents an innovative methodology that leverages ResNet-50, a Convolutional Neural Network (CNN) model, to identify distinct water level features in Closed-Circuit Television (CCTV) river imagery of the Chengmei Bridge on the Keelung River in Neihu District, Taiwan, under various weather conditions. This methodology creates a virtual water gauge system for the precise and timely detection of water levels, thereby eliminating the need for dependable physical gauges. Our study utilized image data from 1 March 2022 to 28 February 2023. This river, crucial to the ecosystems and economies of numerous cities, could instigate a range of consequences due to rapid increases in water levels. The proposed system integrates grid-based methods with infrastructure like CCTV cameras and Raspberry Pi devices for data processing. This integration facilitates real-time water level monitoring, even without physical gauges, thus reducing deployment costs. Preliminary results indicate an accuracy range of 83.6% to 96%, with clear days providing the highest accuracy and heavy rainfall the lowest. Future work will refine the model to boost accuracy during rainy conditions. This research introduces a promising real-time river water level monitoring solution, significantly contributing to flood control and disaster management strategies.補正完畢CH
PIFTA4Rec: Leveraging Personalized Item Frequency and Temporal Attention for Enhanced Next-Basket Recommendation
No full textIn the digital age, recommendation systems play a vital role in alleviating information overload, enhancing user engagement, and driving significant growth in e-commerce. While widely adopted in next-basket prediction tasks, conventional methods primarily focus on short-term user interactions and often overlook long-term behavioral patterns that are crucial for delivering personalized recommendations. To address this limitation, we propose a novel approach: Personalized Item Frequency (PIF), a key feature that models users’ repeated purchase behaviors over time. Integrating PIF allows for the capture of subtle and consistent buying habits, thereby improving recommendation accuracy beyond traditional frequency-based or recency-oriented models. Building on this foundation, we introduce PIFTA4Rec, a hybrid neural network model designed to enhance both recommendation precision and computational efficiency. PIFTA4Rec combines K-Nearest Neighbor (KNN) for PIF-based vector prediction with a Temporal Attention mechanism to accurately model the timing of user purchases. In addition, the model leverages the multi-head attention mechanism of Transformers to capture complex and dynamic user–item relationships across diverse contexts, making it particularly effective for next-basket recommendation scenarios. Empirical evaluations on two real-world datasets demonstrate that PIFTA4Rec consistently outperforms state-of-the-art next-basket recommendation models in terms of both accuracy and robustness. These results underscore the importance of incorporating long-term purchase patterns—such as those captured by PIF—in advancing next-basket recommendation systems. This study introduces a unified and interpretable next-basket recommendation framework by integrating traditional PIF-based modeling with deep learning techniques, delivering both theoretical insights and practical benefits for future research and applications.補正完畢US
Microwave Imaging of Uniaxial Objects Using a Hybrid Input U-Net
No full textThis paper introduces hybrid inputs using Internet of Things (IoT) sensors for reconstructing microwave images of uniaxial objects. Specifically, scattered field data is obtained through IoT sensors, and artificial intelligence techniques are employed to enable real-time electromagnetic imaging. The presented method combines a U-Net architecture with an integrated input to reconstruct high-resolution images of dielectric targets for both Transverse Magnetic (TM) and Transverse Electric (TE) waves. The z-axial dielectric constants are reconstructed by the TM wave illumination, while the x- and y-axial dielectric constants are recovered by the TE wave illumination. First, a Direct Sampling Method (DSM) gives spatial details of the target. Second, a Back-propagation (BP) scheme provides basic information about the target’s properties. Lastly, we combine these two inputs by taking their product, which is further processed in the U-Net. Numerical results show that this integration can improve image quality with nearly no additional computing burden. Experiments also reveal that our proposed method is both accurate and efficient for uniaxial objects, making it a reliable solution to overcome the challenges in electromagnetic imaging.補正完畢CH
Ultrafiltration membrane fabricated from polyethylene terephthalate plastic waste for treating microalgal wastewater and reusing for microalgal cultivation
No full textCurrent study had made a significant progress in microalgal wastewater treatment through the implementation of an economically viable polyethylene terephthalate (PET) membrane derived from plastic bottle waste. The membrane exhibited an exceptional pure water flux of 156.5 ± 0.25 L/m2h and a wastewater flux of 15.37 ± 0.02 L/m2h. Moreover, the membrane demonstrated remarkable efficiency in selectively removing a wide range of residual parameters, achieving rejection rates up to 99%. The reutilization of treated wastewater to grow microalgae had resulted in a marginal decrease in microalgal density, from 10.01 ± 0.48 to 9.26 ± 0.66 g/g. However, this decline was overshadowed by a notable enhancement in lipid production with level rising from 181.35 ± 0.42 to 225.01 ± 0.11 mg/g. These findings signified the membrane's capacity to preserve nutrients availability within the wastewater; thus, positively influencing the lipid synthesis and accumulation within microalgal cells. Moreover, the membrane's comprehensive analysis of cross-sectional and surface topographies revealed the presence of macropores with a highly interconnected framework, significantly amplifying the available surface area for fluid flow. This exceptional structural attribute had substantially contributed to the membrane's efficacy by facilitating superior filtration and separation process. Additionally, the identified functional groups within the membrane aligned consistently with those commonly found in PET polymer, confirming the membrane's compatibility and efficacy in microalgal wastewater treatment.補正完畢GB
Transverse Electric Inverse Scattering of Conductors Using Artificial Intelligence
No full textSensors are devices that can detect changes in the external environment and convert them into signals. They are widely used in fields like industrial automation, smart homes, medical devices, automotive electronics, and the Internet of Things (IoT), enabling real-time data collection to enhance system intelligence and efficiency. With advancements in technology, sensors are evolving toward miniaturization, high sensitivity, and multifunctional integration. This paper employs the Direct Sampling Method (DSM) and neural networks to reconstruct the shape of perfect electric conductors from the sensed electromagnetic field. Transverse electric (TE) electromagnetic waves are transmitted to illuminate the conductor. The scattered fields in the x- and y-directions are measured by sensors and used in the method of moments for forward scattering calculations, followed by the DSM for initial shape reconstruction. The preliminary shape data obtained from the DSM are then fed into a U-net for further training. Since the training parameters of deep learning significantly affect the reconstruction results, extensive tests are conducted to determine optimal parameters. Finally, the trained neural network model is used to reconstruct TE images based on the scattered fields in the x- and y-directions. Owing to the intrinsic strong nonlinearity in TE waves, different regularization factors are applied to improve imaging quality and reduce reconstruction errors after integrating the neural network. Numerical results show that compared to using the DSM alone, combining the DSM with a neural network enables the generation of high-resolution images with enhanced efficiency and superior generalization capability. In addition, the error rate has decreased to below 15%.補正完畢CH
Convolutional Neural Network-Based Electromagnetic Imaging of Uniaxial Objects in a Half-Space
No full textIn this paper, we adopt artificial intelligence (AI) technology for the electromagnetic imaging of uniaxial objects buried in a half-space environment. The limited measurement angle inherent to half-space configurations significantly increases the difficulty of data collection. This paper discusses the simultaneous emission of Transverse Magnetic (TM) and Transverse Electric (TE) electromagnetic waves to illuminate a uniaxial object embedded in a half-space. The dominant current scheme (DCS) and the backpropagation scheme (BPS) are subsequently employed to compute the initial permittivity distribution, which is then used as a dataset for training Convolutional Neural Networks (CNNs). The numerical results compare the reconstruction capabilities of both methods under identical conditions, demonstrating that the DCS exhibits superior generalization and noise immunity compared to the BPS. These findings confirm the effectiveness of both schemes in reconstructing the dielectric constant distribution of uniaxial objects buried in a half-space.補正完畢CH
SMRT: Surveillance Monitoring and Recognition Techniques for Analyzing Service Behavior in Blurred and Unsteady Video
No full textWith the rapid development of the service industry and increasing customer expectations, traditional mystery shopper audit methods face several challenges, such as time-consuming manual analysis, significant subjective bias, and difficulty in accurately quantifying complex service behaviors. To overcome these limitations, this paper introduces an innovative approach called Surveillance Monitoring and Recognition Techniques (SMRT) for analyzing service behavior. The proposed SMRT achieves precise classification of service behaviors through a two-phase process: coarse-grained and fine-grained analysis. In the coarse-grained phase, the proposed SMRT preprocesses blurred video to extract and emphasize relevant external features, specifically detecting and capturing ‘person’ objects in video frames, thereby effectively filtering out irrelevant frames and reducing computational load. In the fine-grained phase, it performs spatiotemporal feature extraction and utilizes Transformer models to conduct a detailed comparison of target behavioral features across video segments. Simulation results demonstrate that the proposed SMRT significantly enhances recognition performance in terms of accuracy, and F1-score compared to existing methods.補正完畢TW
Exploring Lithium Storage Mechanism and Cycling Stability of Bi2Mo3O12 Binary Metal Oxide Anode Composited with Ti3C2 MXene
No full textMetal oxides are widely evaluated as high-capacity anode candidates for practical lithium ion battery applications, owing to their attractive volumetric and gravimetric capacities compared with the traditional graphite anode. Synergistic effects on improving electrochemical performance of binary metal oxide anodes have been increasingly reported via different working mechanisms for lithium storage instead of simple combination of two single components. Herein, we report on exploring lithium storage mechanism in Bi2Mo3O12 binary metal oxide for the first time as an anode material. In-situ synchrotron X-ray diffraction measurements are performed on this exotic material to elucidate lithium storage behaviors, coupled with voltage-resolved cyclic voltammetry and ex-situ X-ray photoelectron spectroscopy analyses. The Bi2Mo3O12 anode undergoes an irreversible initial conversion reaction, resulting in metallic Bi and Li2MoO4 components through electrochemical lithiation. During successive cycling, these two components reversibly uptake and release Li ions through alloying/de-alloying and intercalation/de-intercalation reactions, by forming corresponding Li3Bi alloy and excessively-lithiated Li2+xMoO4 derivative, respectively. Cycling stability of the Bi2Mo3O12 anode material is considerably enhanced by in-situ composition with Ti3C2-based MXene nanosheets. The Bi2Mo3O12@Ti3C2 composite anode material can deliver an initial charge capacity of approximately 846 mAh g−1 at 50 mA g−1 and retain at 227 mAh g−1 upon prolonged 1000 cycles at 2.5 A g−1 high charge/discharge current density. This work offers some insights into lithium storage mechanism and composite nanostructure design in Bi−Mo−O binary metal oxide anode towards enhanced electrochemical performance.補正完畢DE