Search

This study presents a colorimetric/electrical dual-sensing system (CEDS) for low-power, high-precision, adaptable, and real-time detection of hydrogen sulfide (H 2 S) gas. The lead acetate/poly(vinyl alcohol) (Pb(Ac) 2 /PVA) nanofiber film was transferred onto a polyethylene terephthalate (PET) flexible substrate by electrospinning to obtain colorimetric/electrical sensors. The CEDS was constructed to simultaneously record both the visual and electrical response of the sensor, and the improved Manhattan segmentation algorithm and deep neural network (DNN) were used as its intelligent algorithmic aids to achieve quantitative exposure to H 2 S. By exploring the mechanism of color change and resistance response of the sensor, a dual-sensitivity mechanism explanation model was proposed to verify that the system, as a dual-mode parallel system, can adequately solve the sensor redundancy problem. The results show that the CEDS can achieve a wide detection range of H 2 S from 0.1-100 ppm and identify the H 2 S concentration in 4 s at the fastest. The sensor can be stabilized for 180 days with excellent selectivity and a low limit of detection (LOD) to 0.1 ppm of H 2 S. In addition, the feasibility of the CEDS for measuring H 2 S levels in underground waterways was validated. This work provides a new method for adaptable, wide range of applications and low-power, high-precision H 2 S gas detection.

In inverse synthetic aperture radar (ISAR) imaging, range alignment (RA) is crucial for translational compensation. To address the need for rapidity and accuracy in the RA process, a fast and robust range alignment method is proposed based on a deep learning network and minimum entropy (ME) method. The proposed method consists primarily of two components: the CNN-RNN attention mechanism network (CRAN) architecture and the regional multi-scale minimum entropy (RMSME) method. The main distinction of this method from existing approaches lies in its utilization of a deep learning network for rapid coarse alignment, followed by the search for minimum entropy within local regions at multiple scales. The integration strategy effectively addresses the current challenges of poor generalization in deep learning networks and low efficiency in the traditional ME method. The experimental results of simulation data indicate that the proposed method achieves the best range alignment performance compared to RNN, CRAN, and the traditional ME method. The experimental results of the measured data further validate the practicality of the proposed method. This research provides reference significance for the joint application of deep learning and traditional methods in the RA process. [ABSTRACT FROM AUTHOR]

Amortized simulation-based neural posterior estimation provides a novel machine learning based approach for solving parameter estimation problems. It has been shown to be computationally efficient and able to handle complex models and data sets. Yet, the available approach cannot handle the in experimental studies ubiquitous case of missing data, and might provide incorrect posterior estimates. In this work, we discuss various ways of encoding missing data and integrate them into the training and inference process. We implement the approaches in the BayesFlow methodology, an amortized estimation framework based on invertible neural networks, and evaluate their performance on multiple test problems. We find that an approach in which the data vector is augmented with binary indicators of presence or absence of values performs the most robustly. Indeed, it improved the performance also for the simpler problem of data sets with variable length. Accordingly, we demonstrate that amortized simulation-based inference approaches are applicable even with missing data, and we provide a guideline for their handling, which is relevant for a broad spectrum of applications. Author summary: In biomedical research, mechanistic models describe dynamic processes, yet inferring their underlying parameters can often be challenging. Bayesian statistics provides an established framework for this by integrating prior knowledge with observed data, and naturally enables uncertainty quantification as a distribution of parameter values is returned. However, classical case-based methods for Bayesian inference can be computationally expensive, particularly when the same model needs to be fitted to different data sets. Recently, deep-learning-based approaches have been developed to streamline the inference procedure, allowing the upfront training cost to amortize when applied to multiple data sets. In this manuscript, we explore approaches to extend the setup to data sets with missing data. In summary, an encoding scheme which exploits data augmentation with binary indicators of presence or absence performs the most robustly across different test problems. [ABSTRACT FROM AUTHOR]

Deep learning methods have greatly promoted the ship detection performance in synthetic aperture radar (SAR) images. However, due to the different imaging mechanisms and target characteristics, detecting ships in SAR imagery precisely and efficiently is still challenging. To address this issue, a streamlined rotational ship detection network (SRSD-Net) is proposed in this letter. First, we construct a lightweight network using one-level feature and a receptive field block (RFB) to extract ship features with rich context. Second, a simple and effective representation for rotated ship bounding boxes using the minimum external horizontal box and two phase factors is proposed. Finally, to reduce the missed detections of small ships, k-nearest label assignment (KNLA) is designed to make all targets equally contribute to the network training. Extensive experiments are conducted on high-resolution SAR images dataset (HRSID) and the results show that the proposed method can achieve high detection accuracy with low computational cost. [ABSTRACT FROM AUTHOR]

This study addresses the complex challenges associated with road traffic flow prediction and congestion management through the enhancement of the attention-based spatiotemporal graph convolutional network (ASTGCN) algorithm. Leveraging toll data and real-time traffic flow information from Orange County, California, the algorithm undergoes refinement to adeptly capture abrupt changes in road traffic dynamics and identify instances of acute congestion. The optimization of the graph structure is approached from both macro and micro perspectives, incorporating key factors such as road toll information, node connectivity, and spatial distances. A novel graph self-learning module is introduced to facilitate real-time adjustments, while an attention mechanism is seamlessly integrated into the spatiotemporal graph convolution module. The resultant model, termed AASTGNet, exhibits superior predictive accuracy compared to existing methodologies, with MAE, RMSE, and MAPE values of 8.6204, 14.0779, and 0.2402, respectively. This study emphasizes the importance of incorporating tolling schemes in road traffic flow prediction, addresses static graph structure limitations, and adapts dynamically to temporal variations and unexpected road events. The findings contribute to advancing the field of traffic prediction and congestion management, providing valuable insights for future research and practical applications. [ABSTRACT FROM AUTHOR]

Chronic obstructive pulmonary disease (COPD), which has a high prevalence and mortality rate, is an irreversible condition marked by airflow restriction with different degrees of reversible damage. Notably, there is no cure for COPD, whose treatment primarily relies on rehabilitation exercises to improve airflow limitation. In this paper, a vision-based rehabilitation exercise efficacy prediction system is proposed to assess the efficacy of rehabilitation training for COPD patients. A camera was utilized to capture rehabilitation training videos of COPD patients, and we also collected various physical indicators. In addition, we used clustering algorithm to divide patients with different rehabilitation effects for subsequent progression analysis. Our model achieved a classification of rehabilitation progress accuracy of 90.6%, making it possible to effectively obtain favorable rehabilitation training results without physician supervision. It was meaningful for helping COPD patients get effective feedback when training alone. [ABSTRACT FROM AUTHOR]

Deep learning has achieved significant success in various synthetic aperture radar (SAR) imagery interpretation tasks. However, automatic aircraft detection is still challenging due to the high labeling cost and limited data quantity. To address this issue, we propose a multi-stage domain adaptation training framework to efficiently transfer the knowledge from optical imagery and boost SAR aircraft detection performance. To overcome the significant domain discrepancy between optical and SAR images, the training process can be divided into three stages: image translation, domain adaptive pretraining, and domain adaptive finetuning. First, CycleGAN is used to translate optical images into SAR-style images and reduce global-level image divergence. Next, we propose multilayer feature alignment to further reduce the local-level feature distribution distance. By applying domain adversarial learning in both the pretrain and finetune stages, the detector can learn to extract domain-invariant features that are beneficial to the learning of generic aircraft characteristics. To evaluate the proposed method, extensive experiments were conducted on a self-built SAR aircraft detection dataset. The results indicate that by using the proposed training framework, the average precision of Faster RCNN gained an increase of 2.4, and that of YOLOv3 was improved by 2.6, which outperformed other domain adaptation methods. By reducing the domain discrepancy between optical and SAR in three progressive stages, the proposed method can effectively mitigate the domain shift, thereby enhancing the efficiency of knowledge transfer. It greatly improves the detection performance of aircraft and offers an effective approach to address the limited training data problem of SAR aircraft detection. [ABSTRACT FROM AUTHOR]

Diabetic Retinopathy (DR) is a major cause of blindness worldwide. Early detection and treatment are crucial to prevent vision loss, making accurate and timely diagnosis critical. Deep learning technology has shown promise in the automated diagnosis of DR, and in particular, multi-lesion segmentation tasks. In this paper, we propose a novel Transformer-based model for DR segmentation that incorporates hyperbolic embeddings and a spatial prior module. The proposed model is primarily built on a traditional Vision Transformer encoder and further enhanced by incorporating a spatial prior module for image convolution and feature continuity, followed by feature interaction processing using the spatial feature injector and extractor. Hyperbolic embeddings are used to classify feature matrices from the model at the pixel level. We evaluated the proposed model's performance on the publicly available datasets and compared it with other widely used DR segmentation models. The results show that our model outperforms these widely used DR segmentation models. The incorporation of hyperbolic embeddings and a spatial prior module into the Vision Transformer-based model significantly improves the accuracy of DR segmentation. The hyperbolic embeddings enable us to better capture the underlying geometric structure of the feature matrices, which is important for accurate segmentation. The spatial prior module improves the continuity of the features and helps to better distinguish between lesions and normal tissues. Overall, our proposed model has potential for clinical use in automated DR diagnosis, improving accuracy and speed of diagnosis. Our study shows that the integration of hyperbolic embeddings and a spatial prior module with a Vision Transformer-based model improves the performance of DR segmentation models. Future research can explore the application of our model to other medical imaging tasks, as well as further optimization and validation in real-world clinical settings. [ABSTRACT FROM AUTHOR]

Gross primary productivity (GPP) is an important indicator in research on carbon cycling in terrestrial ecosystems. High-accuracy GPP prediction is crucial for ecosystem health and climate change assessments. We developed a site-level GPP prediction method based on the GeoMAN model, which was able to extract spatiotemporal features and fuse external environmental factors to predict GPP on the Tibetan Plateau. We evaluated four models' behavior—Random Forest (RF), Support Vector Machine (SVM), Deep Belief Network (DBN), and GeoMAN—in predicting GPP at nine flux observation sites on the Tibetan Plateau. The GeoMAN model achieved the best results (R2 = 0.870, RMSE = 0.788 g Cm−2 d−1, MAE = 0.440 g Cm−2 d−1). Distance and vegetation type of the flux sites influenced GPP prediction, with the latter being more significant. The different grassland vegetation types exhibited different sensitivity to environmental factors (Ta, PAR, EVI, NDVI, and LSWI) for GPP prediction. Among them, the site located in the alpine swamp meadow was insensitive to changes in environmental factors; the GPP prediction accuracy of the site located in the alpine meadow steppe decreased significantly with the changes in environmental factors; and the GPP prediction accuracy of the site located in the alpine Kobresia meadow also varied with environmental factor changes, but to a lesser extent than the former. This study provides a good reference that deep learning model is able to achieve good accuracy in GPP simulation when considers spatial, temporal, and environmental factors, and the judgement made by deep learning model conforms to basic knowledge in the relevant field. [ABSTRACT FROM AUTHOR]

Construction tool detection is an important link in the operation and maintenance management of professional facilities in public works. Due to the large number and types of construction equipment and the complex and changeable construction environment, manual checking and inventory are still required. It is very challenging to count the variety of tools in a full-time environment automatically. To solve this problem, this paper aims to develop a full-time domain target detection system based on a deep learning network for difficult, complex railway environment image recognition. First, for the different time domain images, the image enhancement network with brightness channel decision is used to set different processing weights according to the images in different time domains to ensure the robustness of image enhancement in the entire time domain. Then, in view of the collected complex environment and the overlapping placement of the construction tools, a lightweight attention module is added on the basis of YOLOX, which makes the detection more purposeful, and the features cover more parts of the object to be recognized to improve the model. Overall detection performance. At the same time, the CIOU loss function is used to consider the distance fully, overlap rate, and penalty between the two detection frames, which is reflected in the final detection results, which can bring more stable target frame regression and further improve the recognition accuracy of the model. Experiments on the railway engineering dataset show that our RYOLO achieves a mAP of 77.26% for multiple tools and a count frame rate of 32.25FPS. Compared with YOLOX, mAP increased by 3.16%, especially the AP of woven bags with a high overlap rate increased from 0.15 to 0.57. Therefore, the target detection system proposed in this paper has better environmental adaptability and higher detection accuracy in complex railway environments, which is of great significance to the development of railway engineering intelligence. [ABSTRACT FROM AUTHOR]

Background: Multiparametric magnetic resonance imaging (MRI) is widely used in breast cancer screening. Accurate prediction of the axillary lymph nodes metastasis (ALNM) is essential for breast cancer surgery and treatment. However, there is no mature and effective discerning method for ALNM based on multiparametric MRI. Purpose: To evaluate the ALNM using T1‐weighted imaging (T1WI), T2‐weighted imaging (T2WI), and diffusion‐weighted imaging (DWI) sequences, respectively, and construct a quantitative ALNM discerning model of integrated multiparametric MRI. Study Type: Retrospective. Population: Three‐hundred forty‐eight breast cancer patients, 163 with ALNM (99.39% females), and 185 without ALNM (100% females). The dataset was randomly divided into the training set (315 cases) and the testing set (33 cases). Field Strength/Sequence: 1.5 T; T1WI (VIBRANT), T2WI (FSE), and DWI (echo planar imaging [EPI]). Assessment: The lesion region of interest images were cropped and sent to a pretrained ResNet50 network. Then, the results of different sequences were sent to a classifier for ensemble learning to construct the ALNM model of multiparametric MRI. Statistical Tests: Performance indicators such as accuracy, the receiver operating characteristic (ROC) curve, and the area under the ROC curve (AUC) were calculated. Student's t‐test, chi‐square test, Fisher's exact test, and Delong test were performed, and P < 0.05 was considered statistically significant. Results: T2WI performed the best among the three sequences, and achieved the accuracy and AUC of 0.933/0.989 in the testing set. Compared to T1WI with the accuracy and AUC of 0.691/0.806, the increase is significant. While compared to DWI with the accuracy and AUC of 0.800/0.910, the improvement is not significant (P = 0.126). After integrating three sequences, the accuracy and AUC improved to 0.970 and 0.996. Data Conclusion: T2WI performed better than DWI and T1WI in discerning ALNM in this breast cancer dataset. The proposed quantitative model of integrated multiparametric MRI could effectively help the ALNM diagnosis. Level of Evidence: 1 Technical Efficacy Stage: 2 [ABSTRACT FROM AUTHOR]

Recently, deep learning has greatly promoted the development of detection methods for ship targets in synthetic aperture radar (SAR) images. However, existing detection networks are mostly based on large-scale models and high-cost computations, which require high-performance computing equipment to realize real-time processing and limit their hardware transplantation to onboard platforms. To address this problem, a lightweight ship detection network via YOLOX-s is proposed in this paper. Firstly, we remove the computationally heavy pyramidal structure and build a streamlined network based on a one-level feature for higher detection efficiency. Secondly, to expand the limited receptive field and enhance the semantic information of a single-feature map, a residual asymmetric dilated convolution (RADC) block is proposed. Through four branches with different dilation rates, the RADC block can help the detector to capture various ships in complex backgrounds. Finally, to tackle the imbalance problem between ships of different scales in the training stage, we put forward a balanced label assignment strategy called center-based uniform matching. To verify the effectiveness of the proposed method, we conduct extensive experiments on the SAR Ship Detection Dataset (SSDD) and High-Resolution SAR Images Dataset (HRSID). The results show that our method can achieve comparable performance to general detection networks with much less computational cost. [ABSTRACT FROM AUTHOR]

Recoverable and humidity‐tolerant colorimetric sensors are extremely difficult to overcome. In this work, a composite hydrogel sensor based on chelation reaction is proposed for the first time to achieve ultra‐fast recovery and ultra‐low detection limit (LOD) of H2S colorimetric sensing. The composite hydrogel sensor is constructed by polyvinyl alcohol/boric acid/Pb2+ (PVA/Bn/Pb) hydrogel sensing layer and polyacrylamide/sodium alginate (PAM/SA) hydrogel adsorption layer. The change from white to brown is achieved in H2S and can be restored when placed in the air. The response/recovery mechanism is explained by the transfer of Pb2+ between hydrogels and the formation of an “Alginate‐Pb2+” with egg‐box structure during the process. The multimodal fusion deep neural network (MF‐DNN) can overcome the error caused by hydrogel aging. Due to the excellent loading and diffusion function provided by the 3D structure of hydrogel, the sensor has a broad detection range of 0.2–100 ppm, a response/recovery time of 10s/32s, and a theoretical LOD of 0.026 ppm. A smartphone‐based H2S sensing system is developed to enable non‐invasive halitosis diagnosis, egg freshness detection, and remote H2S monitoring and alarming. The system is recoverable, humidity‐tolerant, capable of ultrafast detection and widely applicable. This work provides inspiration for humidity‐tolerant and recoverable H2S gas sensing. [ABSTRACT FROM AUTHOR]

• A novel graph approach to semantic enrichment of BIM models - representing models as graphs and applying GNNs - is presented. • An improved GNN algorithm, SAGE-E, is constructed to process both edge and node features and accelerate computation speed. • SAGE-E achieves higher accuracy and more balanced prediction when compared with other machine learning algorithms. • The effects of features and training sizes are explored. Semantic enrichment of Building Information Modeling (BIM) models supplements models with the implicit semantics for further applications. In this paper, we use the room classification task to develop, test and illustrate a novel approach to semantic enrichment of BIM models - representation of models as graphs and application of graph neural networks (GNNs). A dedicated graph dataset consisting of 224 apartment layouts with nine room types and node/edge features was compiled. An improved GNN algorithm, SAGE-E, was developed for processing both node and edge features and a batch method was used to improve efficiency. The experiments showed that (1) The novel approach of adopting graphs and GNNs was feasible. (2) SAGE-E achieved higher accuracy (79%) and more balanced prediction (F1 = 0.79) when compared with other machine learning algorithms. (3) SAGE-E shortened the training and validation process. This work 1 1 A part of EU CBIM project at https://cbim2020.net.technion.ac.il/ (accessed on 17/8/2021). pioneers the application of GNNs for semantic enrichment and opens the door to other possible applications. [ABSTRACT FROM AUTHOR]

The existing deep learning-based Personal Protective Equipment (PPE) detectors can only detect limited types of PPE and their performance needs to be improved, particularly for their deployment on real construction sites. This paper introduces an approach to train and evaluate eight deep learning detectors, for real application purposes, based on You Only Look Once (YOLO) architectures for six classes, including helmets with four colours, person, and vest. Meanwhile, a dedicated high-quality dataset, CHV, consisting of 1330 images, is constructed by considering real construction site background, different gestures, varied angles and distances, and multi PPE classes. The comparison result among the eight models shows that YOLO v5x has the best mAP (86.55%), and YOLO v5s has the fastest speed (52 FPS) on GPU. The detection accuracy of helmet classes on blurred faces decreases by 7%, while there is no effect on other person and vest classes. And the proposed detectors trained on the CHV dataset have a superior performance compared to other deep learning approaches on the same datasets. The novel multiclass CHV dataset is open for public use. [ABSTRACT FROM AUTHOR]

• Automated classification of BIM objects can improve interoperability. • BIM objects for performance simulation are difficult to classify. • Ensemble learning applies multiple classification tools with different AI methods. • We compiled a dataset with 40 classes and 3,410 performance domain objects. • The best of eight ensemble models tested achieved 91 % accuracy and 0.88 F1 score. Numerous factors in current workflows reduce the completeness, accuracy, and reliability of the information exchanged between Building Information Modelling (BIM) authoring tools and building performance simulation and analysis software. Automated classification of BIM objects could ameliorate most of the interoperability problems. However, architectural, structural, mechanical, electrical, and plumbing (MEP), and other building objects have a wide range of shapes, features, and properties, which makes identifying them reliably and accurately with any single AI tool difficult. This study explored methods and techniques for automated classification of architectural and MEP BIM objects that can significantly reduce or eliminate the manual effort required to correct data exchange and interoperability issues in pre-processing building models for performance evaluation. A series of ensemble models using a stacking mechanism of four classification tools (semantic keyword search, rule-based inferencing, machine learning using object geometry features, and deep learning using visual shape features) were implemented and tested. A BIM object dataset with 3,410 instances belonging to 40 classes was compiled for training and testing the models. The dataset contains diverse architecture and MEP objects with high-quality selected instances and is available for public use. We found that an ensemble learning approach that exploits the various types of information available in instance models may exhibit superior object classification performance than any individual tool. The ensemble model applying all four tools achieved 91.0 % accuracy (F1 score 0.88), suggesting that automated classification using ensemble learning is an effective strategy. [ABSTRACT FROM AUTHOR]

Summary: Motor imagery (MI) is an important control paradigm in the field of brain‐computer interface (BCI), which enables the recognition of personal intention. So far, numerous methods have been designed to classify EEG signal features for MI task. However, deep neural networks have been seldom applied to analyze EEG signals. In this study, two novel kinds of deep learning schemes based on convolutional neural networks (CNN) and Long Short‐Term Memory (LSTM) were proposed for MI‐classification. The frequency domain representations of EEG signals were obtained using short time Fourier transform (STFT) to train models. Classification results were compared between conventional algorithm, CNN, and LSTM models. Compared with two other methods, CNN algorithms had shown better performance. These conclusions verified that CNN method was promising for MI‐based BCIs. [ABSTRACT FROM AUTHOR]

The super-resolution of a very low-resolution face image is a challenge task in single image super-resolution. Most of deep learning methods learn a non-linear mapping of input-to-target space by one-step upsampling. These methods are difficult to reconstruct a high-resolution face image from single very low-resolution face image. In this paper, we propose an asymptotic Residual Back-Projection Network (RBPNet) to gradually learn residual between the reconstructed face image and the ground truth by multi-step residual learning. Firstly, the reconstructed high-resolution feature map is projected to the original low-resolution feature space to generate low-resolution feature map (the projected low-resolution feature map). Secondly, the projected low-resolution feature map is subtracted by original feature map to generate low-resolution residual feature map. And finally, the low-resolution residual feature map is mapped to high-resolution feature space. The network will get a more accurate high-resolution image by iterative residual learning. Meanwhile, we explicitly reconstruct the edge map of face image and embed it into the reconstruction of high-resolution face image to reduce distortion of super-resolution results. Extensive experiments demonstrate the effectiveness and advantages of our proposed RBPNet qualitatively and quantitatively. [ABSTRACT FROM AUTHOR]