IEEE Transactions on Instrumentation and Measurement - new TOC TOC Alert for Publication# 19
- Prediction of Remaining Useful Life of Rolling Bearings Based on Multiscale Efficient Channel Attention CNN and Bidirectional GRUel febrero 20, 2024 a las 1:15 pm
To effectively capture both local and global features while retaining temporal dependencies in time-series data and to improve the accuracy of remaining useful life (RUL) prediction of rolling bearings, this article proposes a hybrid architecture based on a multiscale efficient channel attention convolutional neural network and bidirectional gated recurrent unit (MSECNN-BIGRU) networks. The method is based on MSECNN-BIGRU. The MSECNN module can use both local and global features by incorporating multiscale features and the efficient channel attention (ECA) mechanism. Considering the superiority of a CNN in processing image data, the Gram angle field theory was applied to translate the 1-D vibration signal into Gram’s angle difference field (GADF) image as the input for the MSECNN model. During the subsequent prediction process, bidirectional GRU (BIGRU) networks were proposed to avoid the one-way GRU model ignoring the influence of the next time series. In the BIGRU, the GRU was applied in both forward and backward directions to fully extract relevant information from the front and back of the sequence data, thereby improving the prediction performance of the model. By combining these modules, the MSECNN-BIGRU model could accurately predict the RUL of rolling bearings. The experimental results showed that the MSECNN-BIGRU model outperformed other classical models, making it a reliable model for predicting the RUL of rolling bearings.
- Modeling and Signal Processing of Bulk Acoustic Wave Passive Wireless Strain Sensorsel febrero 19, 2024 a las 1:17 pm
Untethered, battery-less, and chip-less passive wireless strain sensors have been widely investigated to overcome the drawbacks of conventional sensors for structure health monitoring of large civil structures. Although the-state-of-the-art passive wireless sensors enable long-range, high-resolution measurements, the signal processing of these sensors is still a challenging task. Passive wireless sensors require an algorithm to capture their resonant frequencies from noisy signals. In this article, we propose an algorithm based on rational polynomial functions to fit the full waveform of bulk acoustic wave (BAW)-based passive wireless strain sensors. We establish an analytical expression for the signal and simplify it based on multiple constrains. Numerical simulations show that the simplified fitting functions can accurately extract the peak frequency of the resonant signal when these constraints are satisfied. The experimental demonstrations confirm that passive wireless sensors utilizing this algorithm achieve a resolution of $4 \mu \varepsilon $ and a refresh rate of 7.5 Hz. In addition, we used the proposed algorithm to realize the vibration frequency measurement of a cantilever beam with a first mode around 4 Hz. The proposed method has high accuracy and moderate speed in extracting the resonance frequency of passive wireless sensors, thus making it possible to realize noncontact measurements of strain changes or vibrations in large civil structures.
- Two-Stage Progressive Underwater Image Enhancementel febrero 19, 2024 a las 1:17 pm
Underwater image enhancement (UIE) is a challenging problem involving various aspects of image degradation, such as color scattering, low contrast, and haziness. In this study, we present a method named two-stage progressive enhancement network (TPENet) to address these issues. We outline the challenges faced in UIE and introduce how TPENet tackles them. TPENet adopts a two-stage network architecture that combines the extensive context learning capabilities of encoders–decoders and the spatial-detail preservation capabilities of the original resolution network. In the first stage, we design a densely fused encoder–decoder subnetwork that focuses on addressing color distortion and low contrast issues. In the second stage, we introduce an original resolution subnetwork (ORSNet) and tackle the haziness problem in underwater images through an image dehazing auxiliary task. To highlight local features and pass them for further enhancement in the next stage, we also introduce a multicolor space supervised attention module. Through extensive experimental results, we validate the outstanding performance, generalization ability, and positive impact on other visual tasks of the proposed method. Additionally, we conduct ablation studies to demonstrate the contributions of key components.
- RGBT Image Fusion Tracking via Sparse Trifurcate Transformer Aggregation Networkel febrero 19, 2024 a las 1:17 pm
Recently have testified the superior tracking ability of Transformer in RGBT tracking for its global and dynamic modeling property. However, these Transformer-based trackers lack attention to the primary feature information and are susceptible to interference from background information. In addition, they often either focus on shared modality information or specific modality information but fail to adequately explore the potential of these two patterns together. To address these issues, a sparse trifurcate Transformer aggregation network is proposed in this article for enhancing tracking robustness. First, a trifurcate tree structure is designed to obtain both modality-shared and modality-specific information, which can learn more powerful feature representations. Second, a sparse attention mechanism is adopted in Transformer to focus on the important features. To fully mine the complementary multimodal information, a confidence-aware aggregation network is designed to generate reliability weights of each mode. Finally, a double-head network is introduced to locate target. Sufficient experimental results on multiple RGBT benchmarks, including GTOT, RGBT210, RGBT234, and LasHeR, verify superior tracking ability against other advanced trackers.
- Optimizing Rectal Cancer Patient Care: Dworak TRG Prediction via Bayesian Evolutionary Fourier-Domain Random Subspace Forestel febrero 19, 2024 a las 1:17 pm
To achieve personalized and optimized treatment for rectal cancer patients, accurate predictions of treatment response based on pretreatment medical images are essential. However, these images often have varying settings. Our study examined the impact of standardizing pixel spacing and slice thickness on predictive accuracy in medical image analysis. Using our custom-built evolutionary random subspace forest (ERSF) algorithm, we investigated how altering these spatial settings affected the accuracy of tumor regression grade (TRG) prediction. We examined 16 different adjustments to spacing settings in computed tomography (CT) images taken from 139 rectal cancer patients. Furthermore, we explored a Bayesian approach within our random forest (RF) algorithm. We utilized the modified data as prior information and employed radiomics data from the same CT images as posterior information. This study revealed that these alterations notably improved both training and validation accuracy, whereas the Bayesian approach enhanced model generalization, indicating a close alignment between training and validation results.