Search

Two-dimensional MXenes materials are expected to be cost-effective and efficient catalysts for hydrogen evolution reaction (HER) due to their tunable surface electronic structure. To optimize its catalytic activity, this study proposes a data-driven framework (CROSST) to generate descriptors through feature engineering and integrate them into a convolutional neural network to predict the HER performance of transition-metal (TM) anchored bis-transition-metal carbon-nitride (TM-M′ 2 M′′CNO 2) catalysts. The STCNN model achieves a coefficient of determination of more than 0.93, showing the ability to predict the Gibbs free energy of hydrogen adsorption (Δ G H) with high accuracy. We explored the microscopic mechanisms of HER activity through density functional theory and machine learning analysis. It was found that TM anchoring altered the charge distribution of the MXenes material, elevated the atomic orbital occupancy of neighboring O sites, and weakened the O adsorption of H, thus affecting the HER activity. These results provide a theoretical basis for the design of high-performance MXenes HER catalysts and contribute new research perspectives. [Display omitted] • Propose a CNN framework using random forest and symbol transformer for optimization. • Examine the impact of transition metal anchoring on MXenes catalytic performance. • MXenes catalytic behavior is controlled by the outermost and anchoring metals. [ABSTRACT FROM AUTHOR]

Photoacoustic imaging (PAI) is a promising transcranial imaging technique. However, the distortion of photoacoustic signals induced by the skull significantly influences its imaging quality. We aimed to use deep learning for removing artifacts in PAI. In this study, we propose a polarized self-attention dense U-Net, termed PSAD-UNet, to correct the distortion and accurately recover imaged objects beneath bone plates. To evaluate the performance of the proposed method, a series of experiments was performed using a custom-built PAI system. The experimental results showed that the proposed PSAD-UNet method could effectively implement transcranial PAI through a one- or two-layer bone plate. Compared with the conventional delay-and-sum and classical U-Net methods, PSAD-UNet can diminish the influence of bone plates and provide high-quality PAI results in terms of structural similarity and peak signal-to-noise ratio. The 3-D experimental results further confirm the feasibility of PSAD-UNet in 3-D transcranial imaging. PSAD-UNet paves the way for implementing transcranial PAI with high imaging accuracy, which reveals broad application prospects in preclinical and clinical fields. [ABSTRACT FROM AUTHOR]

Switchgear is a vital component of modern power systems, responsible for regulating the flow of electrical power. Contact issues, irregular loads, and other similar problems can cause switchgear to overheat, leading to unexpected disturbances and potential damage to the power equipment. Thermal imaging shows significant potential and is increasingly employed to detect faults in power equipment. However, the unique characteristics of thermal images often pose challenges to accurate fault detection. This research aims to study the effectiveness of transfer learning architectures for switchgear fault detection. This paper applies eleven transfer learning architectures, including SqueezeNet, GoogLeNet, InceptionV3, DenseNet201, ResNet50, Xception, InceptionResNetV2, ShuffleNet, EfficientNetB0, AlexNet, and VGG19. The results of the testing phase demonstrated that the application of transfer learning by fine-tuning pre-trained convolutional neural network architectures was highly effective in the classification of thermal images captured from switchgear units. The models achieved accuracy rates between 83.87% and 98.38%, and values of F 1 -Scores between 83.11% and 98.34% in the pre-trained architectures. [Display omitted] • Transfer learning is applied to address the unique characteristics of thermal images for switchgear fault detection. • A thermal image dataset is setup and augmented to improve the classification performance. • Establish standard comparisons among eleven transfer learning architectures. • Achieve rapid and precise classification with the proposed transfer learning architectures. [ABSTRACT FROM AUTHOR]

• Integration of AI to boost the analysis of defects of architectural coverings layers. • The synergy of AI enhances the sound analysis of customized hardware called PICUS. • A multi-class classification accuracy up to 85% in identifying defective areas. • Up to 99% in binary classification, demonstrating the highest precision. • The method revolutionizes the approach to detection and analysis of defective areas. This paper aims is to show an automated intelligent measurement system for the detection of adhesion defects between architectural antique plaster layers. The method emulates the traditional conservators' procedure based on acoustical perturbations, auscultation, detection and classification. The system makes use of a hardware device, known in literature as PICUS, for the generation and acquisition of acoustic signals, while the processing of the acquired signals is handled by a deep learning (DL) architecture designed ad hoc. After a brief description of the PICUS system and the acoustic data acquisition procedure, the whole architecture of the DL system is carefully described. The proposed method has been validated by a significant case study. The system shows an accuracy of up to 82% (± 2%) in multi-class classification and up to 99% (± 1%) in binary classification. In particular, the obtained results suggest a satisfactory precision in the detection of areas where stabilization is necessary. [ABSTRACT FROM AUTHOR]

Bone diseases deteriorate the microstructure of bone tissue. Optical-resolution photoacoustic microscopy (OR-PAM) enables high spatial resolution of imaging bone tissues. However, the spatiotemporal trade-off limits the application of OR-PAM. The purpose of this study was to improve the quality of OR-PAM images without sacrificing temporal resolution. In this study, we proposed the Photoacoustic Dense Attention U-Net (PADA U-Net) model, which was used for reconstructing full-scanning images from under-sampled images. Thereby, this approach breaks the trade-off between imaging speed and spatial resolution. The proposed method was validated on resolution test targets and bovine cancellous bone samples to demonstrate the capability of PADA U-Net in recovering full-scanning images from under-sampled OR-PAM images. With a down-sampling ratio of [4, 1], compared to bilinear interpolation, the Peak Signal-to-Noise Ratio and Structural Similarity Index Measure values (averaged over the test set of bovine cancellous bone) of the PADA U-Net were improved by 2.325 dB and 0.117, respectively. The results demonstrate that the PADA U-Net model reconstructed the OR-PAM images well with different levels of sparsity. Our proposed method can further facilitate early diagnosis and treatment of bone diseases using OR-PAM. [ABSTRACT FROM AUTHOR]

State-of-the-art models in semantic segmentation primarily operate on single, static images, generating corresponding segmentation masks. This one-shot approach leaves little room for error correction, as the models lack the capability to integrate multiple observations for enhanced accuracy. Inspired by work on semantic change detection, we address this limitation by introducing a methodology that leverages a sequence of observables generated for each static input image. By adding this "temporal" dimension, we exploit strong signal correlations between successive observations in the sequence to reduce error rates. Our framework, dubbed SSG2 (Semantic Segmentation Generation 2), employs a dual-encoder, single-decoder base network augmented with a sequence model. The base model learns to predict the set intersection, union, and difference of labels from dual-input images. Given a fixed target input image and a set of support images, the sequence model builds the predicted mask of the target by synthesizing the partial views from each sequence step and filtering out noise. We evaluate SSG2 across four diverse datasets: UrbanMonitor, featuring orthoimage tiles from Darwin, Australia with four spectral bands at 0.2 m spatial resolution and a surface model; ISPRS Potsdam, which includes true orthophoto images with multiple spectral bands and a 5 cm ground sampling distance; ISPRS Vahingen, which also includes true orthophoto images and a 9 cm ground sampling distance; and ISIC2018, a medical dataset focused on skin lesion segmentation, particularly melanoma. The SSG2 model demonstrates rapid convergence within the first few tens of epochs and significantly outperforms UNet-like baseline models with the same number of gradient updates. However, the addition of the temporal dimension results in an increased memory footprint. While this could be a limitation, it is offset by the advent of higher-memory GPUs and coding optimizations. Our code is available at https://github.com/feevos/ssg2. [ABSTRACT FROM AUTHOR]

• Automating wildlife monitoring with wireless 4G cameras and end-to-end data streams. • Remote monitoring of sensor performance, API handling and automated task management. • Deep learning for automated identification of focal species and human detection. • Total cost saving of >40 % through automation, AI and less regular site visits. • Enabling technologies allow scaling-up of a cost-efficient biodiversity monitoring. Modern approaches with advanced technology can automate and expand the extent and resolution of biodiversity monitoring. We present the development of an innovative system for automated wildlife monitoring in a coastal Natura 2000 nature reserve of the Netherlands with 65 wireless 4G wildlife cameras which are deployed autonomously in the field with 12 V/2A solar panels, i.e. without the need to replace batteries or manually retrieve SD cards. The cameras transmit images automatically (through a mobile network) to a sensor portal, which contains a PostgreSQL database and functionalities for automated task scheduling and data management, allowing scientists and site managers via a web interface to view images and remotely monitor sensor performance (e.g. number of uploaded files, battery status and SD card storage of cameras). The camera trap sampling design combines a grid-based sampling stratified by major habitats with the camera placement along a traditional monitoring route, and with an experimental set-up inside and outside large herbivore exclosures. This provides opportunities for studying the distribution, habitat use, activity, phenology, population structure and community composition of wildlife species and allows comparison of traditional with novel monitoring approaches. Images are transferred via application programming interfaces to external services for automated species identification and long-term data storage. A deep learning model for species identification was tested and showed promising results for identifying focal species. Furthermore, a detailed cost analysis revealed that establishment costs of the automated system are higher but the annual operating costs much lower than those for traditional camera trapping, resulting in the automated system being >40 % more cost-efficient. The developed end-to-end data pipeline demonstrates that continuous monitoring with automated wildlife camera networks is feasible and cost-efficient, with multiple benefits for extending the current monitoring methods. The system can be applied in open habitats of other nature reserves with mobile network coverage. [ABSTRACT FROM AUTHOR]

Deep brain stimulation (DBS) of the subthalamic nucleus (STN) is an effective therapy in ameliorating the motor symptoms of Parkinson disease. However, postoperative optimal contact selection is crucial for achieving the best outcome of deep brain stimulation of the subthalamic nucleus surgery, but the process is currently a trial-and-error and time-consuming procedure that relies heavily on surgeons' clinical experience. In this study, we propose a structural brain connectivity guided optimal contact selection method for deep brain stimulation of the subthalamic nucleus. Firstly, we reconstruct the DBS electrode location and estimate the stimulation range using volume of tissue activated from each DBS contact. Then, we extract the structural connectivity features by concatenating fractional anisotropy and the number of streamlines features of activated regions and the whole brain regions. Finally, we use a convolutional neural network with convolutional block attention module to identify the structural connectivity features for the optimal contact selection. We review the data of 800 contacts from 100 patients with Parkinson disease for the experiment. The proposed method achieves promising results, with the average accuracy of 97.63%, average precision of 94.50%, average recall of 94.46%, and average specificity of 98.18%, respectively. Our method can provide the suggestion for optimal contact selection. Our proposed method can improve the efficiency and accuracy of DBS optimal contact selection, reduce the dependence on surgeons' experience, and has the potential to facilitate the development of advanced DBS technology. [ABSTRACT FROM AUTHOR]

This study aimed to design an autodelineation model based on convolutional neural networks for generating high-risk clinical target volumes and organs at risk in image-guided adaptive brachytherapy for cervical cancer. A novel SERes-u-net was trained and tested using CT scans from 98 patients with locally advanced cervical cancer who underwent image-guided adaptive brachytherapy. The Dice similarity coefficient, 95th percentile Hausdorff distance, and clinical assessment were used for evaluation. The mean Dice similarity coefficients of our model were 80.8%, 91.9%, 85.2%, 60.4%, and 82.8% for the high-risk clinical target volumes, bladder, rectum, sigmoid, and bowel loops, respectively. The corresponding 95th percentile Hausdorff distances were 5.23 mm, 4.75 mm, 4.06 mm, 30.0 mm, and 20.5 mm. The evaluation results revealed that 99.3% of the convolutional neural networks-generated high-risk clinical target volumes slices were acceptable for oncologist A and 100% for oncologist B. Most segmentations of the organs at risk were clinically acceptable, except for the 25% sigmoid, which required significant revision in the opinion of oncologist A. There was a significant difference in the clinical evaluation of convolutional neural networks-generated high-risk clinical target volumes between the two oncologists (P < 0.001), whereas the score differences of the organs at risk were not significant between the two oncologists. In the consistency evaluation, a large discrepancy was observed between senior and junior clinicians. About 40% of SERes-u-net-generated contours were thought to be better by junior clinicians. The high-risk clinical target volumes and organs at risk of cervical cancer generated by the proposed convolutional neural networks model can be used clinically, potentially improving segmentation consistency and efficiency of contouring in image-guided adaptive brachytherapy workflow. [ABSTRACT FROM AUTHOR]

[Display omitted] • The suspicious ultrasonography features are related to immune response in thyroid cancer. • The features were also related to epithelial to mesenchymal transition (EMT). • Immune response and EMT were related to aggressive clinical behaviors. • Immune response and EMT reflected in the decision by radiologists and convolutional neural networks. Ultrasonography (US) features of papillary thyroid cancers (PTCs) are used to select nodules for biopsy due to their association with tumor behavior. However, the molecular biological mechanisms that lead to the characteristic US features of PTCs are largely unknown. This study aimed to investigate the molecular biological mechanisms behind US features assessed by radiologists and three convolutional neural networks (CNN) through transcriptome analysis. Transcriptome data from 273 PTC tissue samples were generated and differentially expressed genes (DEGs) were identified according to US feature. Pathway enrichment analyses were also conducted by gene set enrichment analysis (GSEA) and ClusterProfiler according to assessments made by radiologists and three CNNs - CNN1 (ResNet50), CNN2 (ResNet101) and CNN3 (VGG16). Signature gene scores for PTCs were calculated by single-sample GSEA (ssGSEA). Individual suspicious US features consistently suggested an upregulation of genes related to immune response and epithelial-mesenchymal transition (EMT). Likewise, PTCs assessed as positive by radiologists and three CNNs showed the coordinate enrichment of similar gene sets with abundant immune and stromal components. However, PTCs assessed as positive by radiologists had the highest number of DEGs, and those assessed as positive by CNN3 had more diverse DEGs and gene sets compared to CNN1 or CNN2. The percentage of PTCs assessed as positive or negative concordantly by radiologists and three CNNs was 85.6% (231/273) and 7.1% (3/273), respectively. US features assessed by radiologists and CNNs revealed molecular biologic features and tumor microenvironment in PTCs. [ABSTRACT FROM AUTHOR]

The explication of Convolutional Neural Networks (CNN) through xAI techniques often poses challenges in interpretation. The inherent complexity of input features, notably pixels extracted from images, engenders complex correlations. Gradient-based methodologies, exemplified by Integrated Gradients (IG), effectively demonstrate the significance of these features. Nevertheless, the conversion of these explanations into images frequently yields considerable noise. Presently, we introduce GAD (Gradient Artificial Distancing) as a supportive framework for gradient-based techniques. Its primary objective is to accentuate influential regions by establishing distinctions between classes. The essence of GAD is to limit the scope of analysis during visualization and, consequently reduce image noise. Empirical investigations involving occluded images have demonstrated that the identified regions through this methodology indeed play a pivotal role in facilitating class differentiation. • We propose GAD method to simplify explanations, providing easy-to-interpret maps. • Our gradient-based method focuses on revealing feature importance in CNNs. • GAD minimizes noise in explanations compared to usual gradient-based techniques. • Class distinction is enhanced thanks to our method. • Empirical results show that GAD locates decisive image areas for classification. [ABSTRACT FROM AUTHOR]

The diagnosis of rotator cuff tears (RCTs) using radiographs alone is clinically challenging; thus, the utility of deep learning algorithms based on convolutional neural networks has been remarkable in the field of medical imaging recognition. We aimed to evaluate the diagnostic performance of artificial intelligence (a deep learning algorithm; a convolutional neural network) to detect and classify RCTs using shoulder radiographs, and compare its diagnostic performance with that of orthopedic surgeons. A total of 1169 plain shoulder anteroposterior radiographs (1 image per shoulder) were included in the total dataset and divided into four groups: intact, small, medium, and large to massive tear groups. The sensitivity, specificity, positive predictive value, negative predictive value, accuracy, and area under the receiver operating curve were measured for the detection of RCTs through binary classification. The average accuracy, recall, precision, and F1-score were divided into four groups by cuff tear size for multiclass classification. The convolutional neural network demonstrated a high performance, with 92% sensitivity, 69% specificity, 86% accuracy, and an area under the receiver operating curve of 0.88 for the detection of RCTs. The average accuracy, recall, precision, and F1-score of the convolutional neural network for classification were 60%, 0.42, 0.49, and 0.45, respectively. The accuracy of the convolutional neural network for the detection and classification of RCTs was significantly better than that of orthopedic surgeons. The convolutional neural network demonstrated the diagnostic ability to detect and classify RCTs using plain shoulder radiographs, and the diagnostic performance exhibited equal to superior accuracy when compared with those of shoulder experts. [ABSTRACT FROM AUTHOR]

Myocardial contrast echocardiography (MCE) plays a crucial role in diagnosing ischemia, infarction, masses and other cardiac conditions. In the realm of MCE image analysis, accurate and consistent myocardial segmentation results are essential for enabling automated analysis of various heart diseases. However, current manual diagnostic methods in MCE suffer from poor repeatability and limited clinical applicability. MCE images often exhibit low quality and high noise due to the instability of ultrasound signals, while interference structures can further disrupt segmentation consistency. To overcome these challenges, we proposed a deep-learning network for the segmentation of MCE. This architecture leverages dilated convolutions to capture high-scale information without sacrificing positional accuracy and modifies multi-head self-attention to enhance global context and ensure consistency, effectively overcoming issues related to low image quality and interference. Furthermore, we also adapted the cascade application of transformers with convolutional neural networks for improved segmentation in MCE. In our experiments, our architecture achieved the best Dice score of 84.35% for standard MCE views compared with that of several state-of-the-art segmentation models. For non-standard views and frames with interfering structures (mass), our models also attained the best Dice scores of 83.33% and 83.97%, respectively. These studies proved that our architecture is of excellent shape consistency and robustness, which allows it to deal with segmentation of various types of MCE. Our relatively precise and consistent myocardial segmentation results provide fundamental conditions for the automated analysis of various heart diseases, with the potential to discover underlying pathological features and reduce healthcare costs. [ABSTRACT FROM AUTHOR]

• Accurate infant age predictions can be made using 20 min resting state EEG from a single channel. • The deep learning age prediction model generalises to two independent datasets from two different clinical sites. • The magnitude of the brain age gap differs between infant groups with different Bayley Scale outcomes. Electroencephalography (EEG) can be used to estimate neonates' biological brain age. Discrepancies between postmenstrual age and brain age, termed the brain age gap, can potentially quantify maturational deviation. Existing brain age EEG models are not well suited to clinical cot-side use for estimating neonates' brain age gap due to their dependency on relatively large data and pre-processing requirements. We trained a deep learning model on resting state EEG data from preterm neonates with normal neurodevelopmental Bayley Scale of Infant and Toddler Development (BSID) outcomes, using substantially reduced data requirements. We subsequently tested this model in two independent datasets from two clinical sites. In both test datasets, using only 20 min of resting-state EEG activity from a single channel, the model generated accurate age predictions: mean absolute error = 1.03 weeks (p-value = 0.0001) and 0.98 weeks (p-value = 0.0001). In one test dataset, where 9-month follow-up BSID outcomes were available, the average neonatal brain age gap in the severe abnormal outcome group was significantly larger than that of the normal outcome group: difference in mean brain age gap = 0.50 weeks (p-value = 0.04). These findings demonstrate that the deep learning model generalises to independent datasets from two clinical sites, and that the model's brain age gap magnitudes differ between neonates with normal and severe abnormal follow-up neurodevelopmental outcomes. The magnitude of neonates' brain age gap, estimated using only 20 min of resting state EEG data from a single channel, can encode information of clinical neurodevelopmental value. [ABSTRACT FROM AUTHOR]

23Na MRI can be used to quantify in-vivo tissue sodium concentration (TSC), but the inherently low 23Na signal leads to long scan times and/or noisy or low-resolution images. Reconstruction algorithms such as compressed sensing (CS) have been proposed to mitigate low signal-to-noise ratio (SNR); although, these can result in unnatural images, suboptimal denoising and long processing times. Recently, machine learning has been increasingly used to denoise 1H MRI acquisitions; however, this approach typically requires large volumes of high-quality training data, which is not readily available for 23Na MRI. Here, we propose using 1H data to train a denoising convolutional neural network (CNN), which we subsequently demonstrate on prospective 23Na images of the calf. 1893 1H fat-saturated transverse slices of the knee from the open-source fastMRI dataset were used to train denoising CNNs for different levels of noise. Synthetic low SNR images were generated by adding gaussian noise to the high-quality 1H k-space data before reconstruction to create paired training data. For prospective testing, 23Na images of the calf were acquired in 10 healthy volunteers with a total of 150 averages over ten minutes, which were used as a reference throughout the study. From this data, images with fewer averages were retrospectively reconstructed using a non-uniform fast Fourier transform (NUFFT) as well as CS, with the NUFFT images subsequently denoised using the trained CNN. CNNs were successfully applied to 23Na images reconstructed with 50, 40 and 30 averages. Muscle and skin apparent TSC quantification from CNN-denoised images were equivalent to those from CS images, with <0.9 mM bias compared to reference values. Estimated SNR was significantly higher in CNN-denoised images compared to NUFFT, CS and reference images. Quantitative edge sharpness was equivalent for all images. For subjective image quality ranking, CNN-denoised images ranked equally best with reference images and significantly better than NUFFT and CS images. Denoising CNNs trained on 1H data can be successfully applied to 23Na images of the calf; thus, allowing scan time to be reduced from ten minutes to two minutes with little impact on image quality or apparent TSC quantification accuracy. [ABSTRACT FROM AUTHOR]

Recently, computer vision and artificial intelligence are being used as enabling technologies for plant phenotyping studies, since they allow the analysis of large amounts of data gathered by the sensors. Plant phenotyping studies can be devoted to the evaluation of complex plant traits either on the aerial part of the plant as well as on the underground part, to extract meaningful information about the growth, development, tolerance, or resistance of the plant itself. All plant traits should be evaluated automatically and quantitatively measured in a non-destructive way. This paper describes a novel approach for identifying plant roots from images of the root system architecture using a convolutional neural network (CNN) that operates on small image patches calculating the probability that the center point of the patch is a root pixel. The underlying idea is that the CNN model should embed as much information as possible about the variability of the patches that can show chaotic and heterogeneous backgrounds. Results on a real dataset demonstrate the feasibility of the proposed approach, as it overcomes the current state of the art. • Root systems must be monitored to assess the growth and well-being of a plant. • State-of-the-art approaches mainly use classic ML or U-networks for segmentation. • CNNs can be used for monitoring considering patch-based information. • These models are simpler and faster, and provide better segmentation performance. [ABSTRACT FROM AUTHOR]

In the past few years, the lane detection technique has become a key factor for autonomous driving systems and self-driving cars on the road. Among the various vehicle subsystems, the lane detection module is one of the essential parts of the Advanced Driver Assistance System (ADAS). Conventional lane detection approaches use machine vision algorithms to find straight lines in road scene images. However, it is challenging to identify straight or curved lane markings in complex environments. To deal with this problem, this paper presents a lane detection technique based on deep learning. It is combined with a 3D convolutional network, so the temporal information is added to the network architecture. Using the front camera images, the system can immediately detect the lane marking information ahead. Moreover, we propose an approach for improvement by adding the time axis to the network architecture. In addition to using 3D-ResNet50, the temporal convolution and spatial convolution are separated for processing. The accuracy is 91.34% improved after adding time and space split convolution with LeakyReLU. The experiments carried out using real scene images have demonstrated the feasibility of the proposed technique for applications to various complex scenes. [ABSTRACT FROM AUTHOR]

• A novel optimization algorithm integrating random and grid search mechanisms as function calls. • A cost-effective low inference time deep learning model for detecting 2D ionospheric plasma structures from all-sky airglow images. • The issue of uncertainty quantification has been addressed using the approach of bootstrapping. The ionosphere, part of the Earth's atmosphere, where different plasma irregularities/structures are generated through various electrodynamical processes. Airglow imaging is one of the tools to observe various morphological features of these irregularities. We aim to help all-sky airglow imaging-based ionospheric research by proposing novel deep learning at the edge framework for the detection of different types of mid-latitude ionospheric plasma structures (medium scale traveling ionospheric disturbances with single and multiple bands, and plasma bubbles). The primary challenge was to find an optimal deep neural network by considering the trade-off between accuracy and inference time. In order to address this, a novel hyperparameter optimization technique has been used that integrates the approaches of random and grid search mechanisms as a compact function. The random function generates the subspace for hyperparameters to check the convergence of the model while the grid search creates possible combinations to tune these hyperparameters. Our novel optimization method for deep learning inference at the edge led us to a low-cost, high-accuracy convolutional neural network (CNN) model, which outperformed the complex state-of-the-art deep learning models such as Inception-v3, DenseNet169, VGG16, and VGG19. For airglow image-based plasma structure detection, the proposed model recorded an accuracy of 99.9% with an inference time of 5.8 s. This was an improvement of about 4% in accuracy while 85% reduction in inference time over the best-performing baseline. There was about 93% reduction in the number of model parameters with respect to the best-performing baseline. Uncertainty quantification has also been performed in the present work through bootstrapping to validate the robustness of the proposed model. The findings of our study show the promise of deep learning at the edge for all-sky airglow imaging systems by demonstrating an alternate low-cost, high-accuracy deep neural network. [ABSTRACT FROM AUTHOR]

Evaluation of left ventricular (LV) function in critical care patients is useful for guidance of therapy and early detection of LV dysfunction, but the tools currently available are too time-consuming. To resolve this issue, we previously proposed a method for the continuous and automatic quantification of global LV function in critical care patients based on the detection and tracking of anatomical landmarks on transesophageal heart ultrasound. In the present study, our aim was to improve the performance of mitral annulus detection in transesophageal echocardiography (TEE). We investigated several state-of-the-art networks for both the detection and tracking of the mitral annulus in TEE. We integrated the networks into a pipeline for automatic assessment of LV function through estimation of the mitral annular plane systolic excursion (MAPSE), called autoMAPSE. TEE recordings from a total of 245 patients were collected from St. Olav's University Hospital and used to train and test the respective networks. We evaluated the agreement between autoMAPSE estimates and manual references annotated by expert echocardiographers in 30 Echolab patients and 50 critical care patients. Furthermore, we proposed a prototype of autoMAPSE for clinical integration and tested it in critical care patients in the intensive care unit. Compared with manual references, we achieved a mean difference of 0.8 (95% limits of agreement: –2.9 to 4.7) mm in Echolab patients, with a feasibility of 85.7%. In critical care patients, we reached a mean difference of 0.6 (95% limits of agreement: –2.3 to 3.5) mm and a feasibility of 88.1%. The clinical prototype of autoMAPSE achieved real-time performance. Automatic quantification of LV function had high feasibility in clinical settings. The agreement with manual references was comparable to inter-observer variability of clinical experts. [ABSTRACT FROM AUTHOR]

High-frequency signals like vibration and acoustic emission are crucial for condition monitoring, but their high sampling rates challenge data acquisition, especially for online monitoring. Our research developed a novel method for condition identification in undersampled signals using a modified convolutional neural network integrated with a signal enhancement approach. A frequency-domain filtering is applied to suppress similar sidebands and obtain more discriminative features of different conditions, followed by an interpolation-based upsampling in the time domain to restore the signal length and strengthen the low-frequency harmonic information. Enhanced signals are converted into two-dimensional grayscale images for neural network analysis. Tested on bearing datasets and real-world data from regenerative thermal oxidizer lift valve leakage, our method effectively extracts features from low-frequency signals, achieving over 95% fault identification accuracy. • This study introduces a novel 2DCNN-based method for condition identification. It reveals a significant accuracy decline when the sampling frequency falls below the Nyquist rate. • A novel signal enhancement approach is presented, suppressing sidebands to enhance feature discrimination in the frequency domain and bolstering low-frequency harmonic information via cubic interpolation in the time domain. • Tested on the CWRU bearing dataset and real-world regenerative thermal oxidizer data, it achieves over 95% accuracy in condition recognition with undersampled signals. [ABSTRACT FROM AUTHOR]

Timely and effective fault diagnosis is crucial to guarantee the reliability and durability of proton exchange membrane fuel cell (PEMFC). Segmented cell technology has been proven to be a powerful tool for exploring the internal characteristics of fuel cells, yet it is seldom applied for fault diagnosis. This paper proposes a data-driven fault diagnosis method that combines segmented cell and deep learning technology for PEMFC water management fault diagnosis. The dynamic characteristics of the current distribution of the fuel cell under different levels of faults are explored with segmented cell technology. A dual-input convolutional neural network is applied to combine current distribution data and sensor data for a more comprehensive system status for fault diagnosis. The results show that the model can efficiently identify flooding and drying faults with more than 98.5% accuracy, faster and more accurately than other algorithms. This work highlights the significance of segmented cell technology for the comprehension of water management inside PEMFC and provides a new concept for PEMFC fault diagnosis. [Display omitted] • The segmented cell technology is applied to investigate the current distribution during flooding and drying process. • A dual-input CNN is adopted to establish a fault diagnosis model for PEMFC water management. • The proposed method exhibits excellent diagnostic performance, with an accuracy rate of over 98.5%. [ABSTRACT FROM AUTHOR]

The large number of near-Earth asteroids (NEAs) has greatly impacted human space activities and Earth security. However, detecting NEAs in astronomical images with complex, varying backgrounds is still extremely challenging. In this paper, we propose a deep segmentation assisted asteroid tracking algorithm, termed DSAT, to construct a possible pipeline for faint NEA tracking in astronomical images. First, the single-frame object detection problem is converted to a segmentation problem, enabling robust extraction of faint potential moving objects. Then, a multiframe motion prior-based moving object tracking algorithm is proposed to find real NEAs. We further propose a distance tolerance criterion to help DSAT achieve effective tracking in practical situations when detection has partially failed. Finally, the pipeline is tested with both simulated and real astronomical images at different SNRs and in crowded fields. The results showed that our pipeline has the potential to detect and track faint NEAs in complex backgrounds. Our code is publicly available at https://github.com/zhenhongdu/DeepSegAsteroidTracker. [ABSTRACT FROM AUTHOR]

Skin lesions refer to changes or abnormalities on the skin, such as moles, freckles, cysts, warts, and more. While not all skin lesions are cancerous, some types can indicate or develop into skin cancer. Skin cancer has been on the rise in recent years, causing many deaths. However, manual examination by dermatologists is expensive and time-consuming. Research in this field has focused on developing an automated system to detect skin cancer and reduce the associated mortality rate. However, the performance of such systems needs improvement. In this study, we propose a Synthetic Minority Over-sampling Technique combined with Edited Nearest Neighbors (SMOTEENN) to augment the data, resulting in a more balanced dataset. Then, we used a Convolutional Neural Network (CNN) model on the benchmark HAM10000 dataset. The dataset comprises seven skin diseases, including Actinic Keratoses, Basal cell carcinoma, Benign keratosis, Dermatofibroma, Melanocytic nevi, Melanoma, and Vascular skin lesions. With a balanced dataset, the proposed model achieves an accuracy, precision, recall, and f1-score of 99.35%, 99.00%, 99.00%, and 99.00%, respectively. The balanced dataset performs 22% better than an unbalanced dataset. [ABSTRACT FROM AUTHOR]

Spatiotemporal fusion can provide remote sensing images with both high temporal and high spatial resolution for earth observation applications. Most of the state-of-the-art models require three or even five images as input, which may lead to difficulties in practical applications due to bad weather or data missing. In this paper, the enhanced cross-paired wavelet based spatiotemporal fusion networks (ECPW-STFN) are proposed to improve the accuracy and quality of the fusions with fewer remote sensing images as inputs. Wavelet transform is introduced into spatiotemporal image fusion to achieve separate training of the high and low frequency components of the image to better extract features of different levels. In addition, a compound loss function containing wavelet loss is proposed to facilitate the preservation of details. Also, an enhancement module with convolutional block attention is put forward to further refine the prediction quality. Experiments were conducted to compare the proposed ECPW-STFN with five state-of-the-art methods on the public CIA and Daxing datasets. The results show ECPW-STFN is better than GAN-STFM which also requires two images, and not inferior to the methods requiring at least three inputs, even exceeds the optimal MLFF-GAN in some cases, proving its great superiority and potential. The code will be available at https://github.com/lixinghua5540/ECPW-STFN. [ABSTRACT FROM AUTHOR]

Imaging-based differentiation between glioblastoma (GB) and brain metastases (BM) remains challenging. Our aim was to evaluate the performance of 3D-convolutional neural networks (CNN) to address this binary classification problem. T1-CE, T2WI, and FLAIR 3D-segmented masks of 307 patients (157 GB and 150 BM) were generated post resampling, co-registration normalization and semi-automated 3D-segmentation and used for internal model development. Subsequent external validation was performed on 59 cases (27 GB and 32 BM) from another institution. Four different mask-sequence combinations were evaluated using area under the curve (AUC), precision, recall and F1-scores. Diagnostic performance of a neuroradiologist and a general radiologist, both without and with the model output available, was also assessed. 3D-model using the T1-CE tumor mask (TM) showed the highest performance [AUC 0.93 (95% CI 0.858–0.995)] on the external test set, followed closely by the model using T1-CE TM and FLAIR mask of peri-tumoral region (PTR) [AUC of 0.91 (95% CI 0.834–0.986)]. Models using T2WI masks showed robust performance on the internal dataset but lower performance on the external set. Both neuroradiologist and general radiologist showed improved performance with model output provided [AUC increased from 0.89 to 0.968 (p = 0.06) and from 0.78 to 0.965 (p = 0.007) respectively], the latter being statistically significant. 3D-CNNs showed robust performance for differentiating GB from BMs, with T1-CE TM, either alone or combined with FLAIR-PTR masks. Availability of model output significantly improved the accuracy of the general radiologist. • 3-dimensional CNN models can differentiate GB from BM with high accuracy. • The best performing models used T1-CE images, either alone or combined with FLAIR images. • The model output significantly improved the accuracy of the general radiologist. [ABSTRACT FROM AUTHOR]

In the field of mobile health, portable dynamic electrocardiogram (ECG) monitoring devices often have a limited number of lead electrodes due to considerations, such as portability and battery life. This situation leads to a contradiction between the demand for standard 12‑lead ECG information and the limited number of leads collected by portable devices. This study introduces a composite ECG vector reconstruction network architecture based on convolutional neural network (CNN) combined with recurrent neural network by using leads I, II, and V2. This network is designed to reconstruct three‑lead ECG signals into 12‑lead ECG signals. A 1D CNN abstracts and extracts features from the spatial domain of the ECG signals, and a bidirectional long short-term memory network analyzes the temporal trends in the signals. Then, the ECG signals are inputted into the model in a multilead, single-channel manner. Under inter-patient conditions, the mean reconstructed Root mean squared error (RMSE) for precordial leads V1, V3, V4, V5, and V6 were 28.7, 17.3, 24.2, 36.5, and 25.5 μV, respectively. The mean overall RMSE and reconstructed Correlation coefficient (CC) were 26.44 μV and 0.9562, respectively. This paper presents a solution and innovative approach for recovering 12‑lead ECG information when only three‑lead information is available. After supplementing with comprehensive leads, we can analyze the cardiac health status more comprehensively across 12 dimensions. [ABSTRACT FROM AUTHOR]

Monitoring body condition score (BCS) is a useful management tool to estimate the energy reserves of an individual cow or a group of cows. The aim of this study was to develop and evaluate the performance of a fully automated 2-dimensional imaging system using a machine learning algorithm to generate real-time BCS for dairy cows. Two separate datasets were used for training and testing. The training dataset included 34,150 manual BCS (MAN_BCS) assigned by 5 experienced veterinarians during 35 visits at 7 dairy farms. Ordinal regression methods and deep learning architecture were used when developing the algorithm. Subsequently, the testing dataset was used to evaluate the developed BCS prediction algorithm on 4 of the participating farms. An experienced human assessor (HA1) visited these farms and performed 8 whole-milking-herd BCS sessions. Each farm was visited twice, allowing for 30 d (±2 d) to pass between visits. The MAN_BCS assigned by HA1 were considered the ground truth data. At the end of the validation study, MAN_BCS were merged with the stored automated BCS (AI_BCS), resulting in a testing dataset of 9,657 single BCS. A total of 3,817 cows in the testing dataset were scored twice 30 d (±2 d) apart, and the change in their BCS (ΔBCS) was calculated. A subset of cows at one farm were scored twice on consecutive days to evaluate the within-observer agreement of both the human assessor and the system. The manual BCS of 2 more assessors (HA2 and HA3) were used to assess the interobserver agreement between humans. Finally, we also collected ultrasound measurements of backfat thickness (BFT) from 111 randomly selected cows with available MAN_BCS and AI_BCS. Using the testing dataset, intra- and interobserver agreement for single BCS and ΔBCS were estimated by calculating the simple percentage agreement (PA) at 3 error levels and the weighted kappa (κ w) for the exact agreement. A Bland-Altman plot was constructed to visualize the systematic and proportional bias. The association between MAN_BCS and AI_BCS and the BFT was assessed with Passing-Bablok regressions. The system had an almost perfect repeatability with a κ w of 0.99. The agreement between MAN_BCS and AI_BCS was substantial, with an overall κ w of 0.69. The overall PA at the exact, ± 0.25-unit, and ± 0.50-unit BCS error range between MAN_BCS and AI_BCS was 44.4%, 84.6%, and 94.8%, respectively, and greater than the PA obtained between HA1 and HA3. The Bland-Altman plot revealed a minimal systematic bias of −0.09 with a proportional bias at the extreme scores. Furthermore, despite the low κ w of 0.20, the overall PA at the exact and ± 0.25-unit of BCS error range between MAN_BCS and AI_BCS regarding the ΔBCS was 45.7 and 88.2%, respectively. A strong linear relationship was observed between BFT and AI_BCS (ρ = 0.75), although weaker than that between BFT and MAN_BCS (ρ = 0.91). The system was able to predict single BCS and ΔBCS with satisfactory accuracy, comparable to that obtained between trained human scorers. [ABSTRACT FROM AUTHOR]

To validate the performance of nnU-Net in segmentation and CNN in classification for liver fibrosis using T1-weighted images. In this prospective study, animal models of liver fibrosis were induced by injecting subcutaneously a mixture of Carbon tetrachloride and olive oil. A total of 99 male Wistar rats were successfully induced and underwent MR scanning with no contrast agent to get T1-weighted images. The regions of interest (ROIs) of the whole liver were delineated layer by layer along the liver edge by 3D Slicer. For segmentation task, all T1-weighted images were randomly divided into training and test cohorts in a ratio of 7:3. For classification, images containing the hepatic maximum diameter of every rat were selected and 80% images of no liver fibrosis (NLF), early liver fibrosis (ELF) and progressive liver fibrosis (PLF) stages were randomly selected for training, while the rest were used for testing. Liver segmentation was performed by the nnU-Net model. The convolutional neural network (CNN) was used for classification task of liver fibrosis stages. The Dice similarity coefficient was used to evaluate the segmentation performance of nnU-Net. Confusion matrix, ROC curve and accuracy were used to show the classification performance of CNN. A total of 2628 images were obtained from 99 Wistar rats by MR scanning. For liver segmentation by nnU-Net, the Dice similarity coefficient in the test set was 0.8477. The accuracies of CNN in staging NLF, ELF and PLF were 0.73, 0.89 and 0.84, respectively. The AUCs were 0.76, 0.88 and 0.79, respectively. The nnU-Net architecture is of high accuracy for liver segmentation and CNN for assessment of liver fibrosis with T1-weighted images. [ABSTRACT FROM AUTHOR]

Human age and gender prediction from facial images has garnered significant attention due to its importance in various applications. Traditional models struggle with large-scale variations in unfiltered images. Convolutional Neural Networks (CNNs) have emerged as effective tools for facial analysis due to their robust performance. This paper presents a novel CNN approach for robust age and gender classification using unconstrained real-world images. The CNN architecture includes convolution, pooling, and fully connected layers for feature extraction, dimension reduction, and mapping to output classes. Adience and UTKFace datasets were utilized, with the best training and testing accuracies achieved using an 80% training and 20% testing data split. Robust image pre-processing and data augmentation techniques were applied to handle dataset variations. The proposed approach outperformed existing methods, achieving age prediction accuracies of 86.42% and 81.96%, and gender prediction accuracies of 97.65% and 96.32% on the Adience and UTKFace datasets, respectively. [ABSTRACT FROM AUTHOR]

• Electrocorticography reveals a two-stage face processing in the human brain. • The human brain shows a "detection-first, recognize-later" face procession pattern. • Convolutional neural networks exhibit brain-like face processing pattern. • This sequential organization of face processing is a computational optimization. Face processing includes two crucial processing levels – face detection and face recognition. However, it remains unclear how human brains organize the two processing levels sequentially. While some studies found that faces are recognized as fast as they are detected, others have reported that faces are detected first, followed by recognition. We discriminated the two processing levels on a fine time scale by combining human intracranial EEG (two females, three males, and three subjects without reported sex information) and representation similarity analysis. Our results demonstrate that the human brain exhibits a "detection-first, recognition-later" pattern during face processing. In addition, we used convolutional neural networks to test the hypothesis that the sequential organization of the two face processing levels in the brain reflects computational optimization. Our findings showed that the networks trained on face recognition also exhibited the "detection-first, recognition-later" pattern. Moreover, this sequential organization mechanism developed gradually during the training of the networks and was observed only for correctly predicted images. These findings collectively support the computational account as to why the brain organizes them in this way. [ABSTRACT FROM AUTHOR]

Nowadays, for controlling crime, surveillance cameras are typically installed in all public places to ensure urban safety and security. However, automating Human Activity Recognition (HAR) using computer vision techniques faces several challenges such as lowlighting, complex spatiotemporal features, clutter backgrounds, and inefficient utilization of surveillance system resources. Existing attempts in HAR designed straightforward networks by analyzing either spatial or motion patterns resulting in limited performance while the dual streams methods are entirely based on Convolutional Neural Networks (CNN) that are inadequate to learning the long-range temporal information for HAR. To overcome the above-mentioned challenges, this paper proposes an optimized dual stream framework for HAR which mainly consists of three steps. First, a shots segmentation module is introduced in the proposed framework to efficiently utilize the surveillance system resources by enhancing the lowlight video stream and then it detects salient video frames that consist of human. This module is trained on our own challenging Lowlight Human Surveillance Dataset (LHSD) which consists of both normal and different levels of lowlighting data to recognize humans in complex uncertain environments. Next, to learn HAR from both contextual and motion information, a dual stream approach is used in the feature extraction. In the first stream, it freezes the learned weights of the backbone Vision Transformer (ViT) B-16 model to select the discriminative contextual information. In the second stream, ViT features are then fused with the intermediate encoder layers of FlowNet2 model for optical flow to extract a robust motion feature vector. Finally, a two stream Parallel Bidirectional Long Short-Term Memory (PBiLSTM) is proposed for sequence learning to capture the global semantics of activities, followed by Dual Stream Multi-Head Attention (DSMHA) with a late fusion strategy to optimize the huge features vector for accurate HAR. To assess the strength of the proposed framework, extensive empirical results are conducted on real-world surveillance scenarios and various benchmark HAR datasets that achieve 78.6285%, 96.0151%, and 98.875% accuracies on HMDB51, UCF101, and YouTube Action, respectively. Our results show that the proposed strategy outperforms State-of-the-Art (SOTA) methods. The proposed framework gives superior performance in HAR, providing accurate and reliable recognition of human activities in surveillance systems. [ABSTRACT FROM AUTHOR]

Exploiting the capability of deep learning-based techniques, this research addresses an important and relevant problem on how to cost effectively detect thyroid cancer. In recent decades, there has been a significant increase in the incidence of thyroid cancer, prompting the need to understand its epidemiology. Existing studies have primarily employed qualitative techniques to investigate a single risk factor correlated with the disease development at a time. However, such an approach is inefficient and tends to ignore the interwoven connections among factors, resulting in a considerable disagreement with the identified risk factors among scholars. Additionally, the use of deep learning techniques in conjunction with medical imaging for computer-aided diagnostic (CAD) systems design has shown promises in detecting the disease, while there are research gaps regarding the detection of subtype and their co-existence. More importantly, existing CAD systems have shortcomings in adapting to different sample groups. To address these challenges, this research aims to shed light on the pathogenesis of thyroid cancer, enhance diagnostic performance, and improve generalisation of deep learning-based decision support systems. By harnessing the power of machine learning, specifically data mining and deep learning techniques, we seek to improve our understanding of the underlying mechanisms of the disease and develop innovative, robust, accurate, and efficient diagnostic tools. Extensive experiments indicate superior performance of the proposed methods than existing works. The systems proposed in this study have great impact to the wider society and contribute to the advancement of thyroid cancer research while enhancing clinical practice in related detection and subsequent management. [ABSTRACT FROM AUTHOR]

Tomato diseases have become a major concern to the tomato production sector around the world. A huge proportion of tomato crops deteriorate yearly during growth or after harvesting due to the infections caused by fungus, viruses and bacteria. Early detection of these diseases plays a crucial role in alleviating overall production loss. Over the past decades, farmers have been using visual observation to identify a crop disease in a field. However, the visual observation method is labour intensive, time consuming, and prone to human error. Currently, intelligent approaches have been widely used to detect and classify these diseases. The objective of this study is to design a convolutional neural network VGG16 net architecture that is able to detect tomato diseases on tomato leaves. A model that can successfully detect 5 tomato leaf diseases was developed: (1) Bacterial spot, (2) Early blight, (3) Late blight, (4) Septoria leaf and (5) Yellow leaf curl with a training accuracy of 0.9328. [ABSTRACT FROM AUTHOR]

The continuous growth of blockchain has brought revolutionary impact on different areas with characteristics like decentralization, trust-worthy and tampering-proof. In the financial field, forecasting stock market had remained a challenging task for researchers. A novel integration of convolution and LSTM along with attention mechanism is proposed for prediction of stock price for the last hour of a day based it's performance in six hours for making prediction in intraday trades. The proposed prediction model uses 4 different stocks of Apple, Facebook, Nike and Uber for data input. The input data is given to the combination of many convolutional and LSTM layers for predicting future value of stocks in short term. Furthermore, the proposed model combines the obtained data with LSTM layers output vectors enhanced for final prediction. The prediction is based on past value of stocks. Performance evaluation is done based on RMSE and computational time. The average values obtained were 0.091 and 11.57 seconds respectively, showing it's superiority from benchmark models. The obtained data is stored on blockchain for enhanced security. The model can further be tuned for predicting other markets and will be more efficient than market specific markets. [ABSTRACT FROM AUTHOR]

The present study uses the benchmark UNSW-NB15 datasets to detect network anomalies using the proposed deep learning-based Convolutional Neural Network (CNN) model. Several studies have already worked on network anomaly detection with some limits. However, earlier research has used old datasets and obtained limited accuracy. In addition, previous studies could not detect the latest malicious activities. Therefore, in the present study, firstly, we implemented the machine learning-based random forest method to diminish the dimensionality of the dataset and pick the most notable features. In this case, the random forest method has set only fifteen features from the forty-one and reduced the complexity of the data. Subsequently, the CNN model was trained and tested on the UNSW-NB15 dataset with 99.00% testing accuracy. The performance of the offered CNN model was assessed using precision, recall and the f1-values. The present study's outcome can be used in real-time malicious activity detection. [ABSTRACT FROM AUTHOR]

Anomalies such as microaneurysms, exudates, and hemorrhages are diagnosed by retinal vessel segmentation to identify different phases of Diabetic Retinopathy (DR); several typical ways to detect Hard Exudates (HE) in retinal images have been used to determine the severity of diabetes. This work proposes a hybrid approach to better segment such abnormal cells. It optimizes the learning parameters of a Convolutional Neural Network (CNN) with a nature-inspired optimizer, the Self-Adaptive Jaya Optimization Algorithm (SAJOA). The SAJOA helps in avoiding trapping in local minima. It uses the SAJOA to steer the search for obtaining near-optimal parameter values for the CNN training. We choose the learning parameters such as batch size, dropout rate, learning rate, etc. The batch size affects the batch normalization performance, and the dropout rate involves the regularization property of the model. The near-optimal values of these two parameters help improve the generalization ability of the deep learning architecture. Thus, this hybrid approach improves the CNN performance and generalizability, leading to precise segmentation of unseen data, which aids in diagnosing diabetic retinopathy. The results of many trials were evaluated against the DRIVE dataset; the proposed method achieved accuracy and F1 score on average, higher by 7 to 8% than the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

In recent days, credit card fraud has emerged as a significant challenge for researchers in the field of detection and prevention. Tackling this challenge holds substantial benefits for both public and private organizations, as it directly impacts economic statistics. Our proposed model introduces crucial advancements to address this issue in real-time scenarios. To achieve this, we harnessed the power of Deep Convolutional Neural Networks (DeepConvNet) in conjunction with various optimization techniques. Optimization algorithms encompass a group of mathematical and computational techniques employed to discover the most suitable solution or set of solutions for a given problem. The primary goal of these algorithms is to optimize, either by maximizing or minimizing, an objective function while ensuring compliance with specific constraints. In this research work, we provide a comparison of highly effective and validated optimization techniques: Stochastic Gradient Descent (Sgd), Adaptive Gradient (Adagrad), Adaptive Moment Estimation (Adam), and Root Mean Squared Propagation (Rmsprop). These optimization algorithms are applied to the Deep Convolutional Neural Network (DeepConvNet) in the subject of our specific problem statement, which involves credit card fraud detection (CCFD). After careful consideration of the problem's nature, objective function characteristics, and computational aspects, we fnd that all four algorithms are suitable for our CCFD task. However, based on experimental results, it is evident that Rmsprop outperforms others, leading to a remarkable 99.93% in accuracy. [ABSTRACT FROM AUTHOR]

Emotion detection has become an important study subject due to the potential uses that it has in many other domains. It incorporates data from various sources, such as facial expressions, voice patterns, physiological indications, and textual signals. To identify human emotions, a real-time bimodal image and audio-based detection-based approach has been proposed. The approach is termed as "bimodal" since images and audio data are considered as inputs to the proposed model. In this study, hybridized CNN-LSTM for the image-based model and Multi-Layer Perceptron is used for the developing audio model. For the image-based model, two combined layers of CNN, and LSTM along with regularization techniques like drop-out, and batch normalization have been adopted for model development. Similarly, for the audio model, the dataset has been fine-tuned using a standard scaler and PCA before training. Based on the combined results of both models, the user's final emotion is determined. Both the audio-based model and the video-based model outperform the existing models discussed in the literature with higher performance accuracy of 79.34% and 77.27% respectively. [ABSTRACT FROM AUTHOR]

The classification of underwater acoustic targets is a challenging task due to the intricacies involved in soundscape, huge amounts of background interference, and the varied features of underwater targets. Various approaches have been explored for classifying underwater targets based on acoustic signatures. The problem with these conventional methods lies in the extensive domain-specific knowledge they demand for effective feature engineering. Alternatively, the deep learning approach is of great appeal for assistance to manual sonar operators who mainly rely on their expertise for the classification of targets. This work leverages the approach of deep neural networks for underwater acoustic target categorization, aiming to analyze whether we can reduce the need for extensive domain knowledge required, by letting deep learning algorithms itself find the deep audio embeddings. This approach implements an audio classifier designed to classify acoustic targets with input as log Mel-spectrograms. To compare the performance of this classifier, we also implement a regular neural network for audio classification with MFCCs as input features. The overall results of the model are promising. [ABSTRACT FROM AUTHOR]

Traffic congestion has become a critical issue in metropolitan areas, inflicting significant hardships on individuals on a daily basis. This adverse situation poses multifaceted challenges, including increased air pollution, heightened fuel consumption, and compromised road safety. Deep learning algorithms, encompassing both Convolutional Neural Networks (CNN), and Recurrent Neural Networks (RNN) have showcased remarkable performance in areas such as computer vision, natural language processing, healthcare, finance, and autonomous systems. This research paper proposes a novel methodology to develop a vehicle detection model - VehiNet, which can subsequently be refined to create an automated traffic signal system. The study explores the effectiveness of various CNN, RNN, and deep learning algorithms and techniques in accomplishing this objective. The dataset picked for the study in addition to being compiled from internet resources, was also painstakingly assembled manually using a variety of methods to collect over 2000 photographs from various angles, lighting conditions, and locations within Bengaluru city. The main motive is to apply CNN, RNN, and other Deep Learning classification techniques for object detection and classification of Indian vehicles and see how well these deep learning algorithms and techniques affect prediction efficiency. This research stands out from others and makes its novelty clear thanks to the carefully collected dataset and the detailed analysis of the methodologies. [ABSTRACT FROM AUTHOR]

The integration of deep learning, computer vision, and advanced algorithms has ushered in a transformative era in the prediction of human driving behavior, consequently revolutionizing road safety. This paper focuses on an innovative convergence of technology that addresses critical issues like driver fatigue and distracted driving by automatically identifying and categorizing diverse driving behaviors. Neural network architectures, such as VGG16, AlexNet, and ResNet are described in this paper that have propelled accuracy in behavior classification to remarkable levels. However, the quest for safer roads remains ongoing, with promising avenues lying ahead. First and foremost, the creation of extensive, diverse, and meticulously annotated datasets is paramount. These datasets serve as the bedrock upon which future models can be trained, enhancing their robustness and generalizability across a spectrum of driving scenarios. Real-time models represent another pivotal frontier. These models hold the potential to provide timely interventions and support systems for drivers, thus preventing accidents proactively. The exploration of hybrid techniques that amalgamate the strengths of various neural network architectures presents an exciting avenue, promising to further push the boundaries of prediction accuracy. Furthermore, this paper also discusses the fusion of multi-modal data, encompassing sensor data from IoT and smartphone devices, that holds immense promise. This holistic approach promises a more comprehensive understanding of driver behavior by integrating diverse data sources, ultimately contributing to the creation of safer road environments.In this research paper, we also explore these cutting-edge developments in deep learning and computer vision, emphasizing technical novelty and innovation. Through an interdisciplinary approach, we envision a future where the synergy of technology, data, and human behavior leads to a substantial reduction in road accidents and improved road safety. [ABSTRACT FROM AUTHOR]

The proliferation of fake images in today's digital landscape poses a significant threat to various domains, including media integrity, social media, and online security. Recognizing the urgent need to distinguish real images from their deceptive counterparts, this paper underscores the importance of developing a robust detection system. While substantial efforts have been made in the realms of computer vision and deep learning, the advent of Generative Adversarial Networks (GANs) has added a new layer of complexity to this challenge. In response to these evolving threats, we present a novel two-step methodology for detecting fake images, with a specific focus on those generated by GANs. Our approach harnesses the combined strengths of GANs and traditional Convolutional Neural Networks (CNNs), offering a comprehensive solution that significantly enhances accuracy in identifying both fake images and machine-generated fake images. The results of our experiments demonstrate the efficacy of our methodology. Using CNNs alone, we achieved a training accuracy of 87%. However, when employing the collaborative power of GANs and CNNs, our model exhibited a remarkable accuracy rate of 94.4%. This substantial improvement underscores the superiority of the GANs+CNN approach, suggesting its potential as a groundbreaking solution in the realm of fake image detection. This research opens up new horizons in fields such as media forensics, social media monitoring, and online security, where the ability to discern genuine content from manipulated or synthetic media is of paramount importance. The promising outcomes of this study not only provide an immediate and effective solution but also pave the way for further exploration and innovation in this critical area of digital security. [ABSTRACT FROM AUTHOR]

The image-based automatic plant disease identification has emerged as a prominent research area, driven by recent advancements in machine learning algorithms. Conventional disease detection methods rely on manual feature extraction from acquired images, involving extensive expertise from experts and pathologists. This labor-intensive practice is often time-consuming and impractical. With the recent advancements in digital technologies and computer vision, Deep neural networks address these challenges by learning the features automatically. This paper focuses on the EfficientNet B0 architecture-based deep neural network model integrated with a convolutional block attention mechanism module for identifying major wheat crop diseases. The attention module in the EfficientNet B0 model improves its performance by emphasizing the disease-infected regions in leaf images. To evaluate the proposed methodology, experiments were conducted on the WheatRust21 image dataset, comprising 6556 images of three major fungal diseases, namely, stripe rust, leaf rust, and stem rust, collected in natural field conditions from 2019 to 2021. Our developed EfficientNet B0-CBAM model performed well with a testing accuracy of 98.70% and an impressive F1 score of 98% on the dataset. The Grad-CAM visualization demonstrates that the attention module focuses more on the infected region and plays a crucial role in wheat disease identification. The results are quantitatively substantial, with the model achieving high accuracy and other evaluation metrics scores with an average of 98.5 %. These findings hold promise for practical applications in agriculture and sustainable crop management. [ABSTRACT FROM AUTHOR]

Major Depressive Disorder (MDD) is characterized by low mood, loss of interest and even suicidal ideation. Electroencephalogram based diagnosis of a variety of neurological conditions has been conducted in recent years using modern neurocomputing and deep learning approaches. Privacy is a high concern for medical data over distributed training scenarios. Hence, this paper aims to develop an EEG-based privacy-preserved system to identify MDD through the Federated Learning (FL) approach. In this study, training for 3-channel EEG-based MDD screening is introduced over FL using deep learning (DL) algorithms such as Long Short-Term Memory (LSTM), Bidirectional LSTM (Bi-LSTM), Gated Recurrent Unit (GRU), and One Dimensional Convolutional Neural Network (1D-CNN). A privacy-preserving solution for MDD patients utilizing FL is advantageous because clinical data is sensitive and most people do not want to share their personal information. Hence, the analysis of the comparative results for multiple applicable strategies including Independent and Identically Disturbed (IID) data, non-IID data, and algorithms provide proof that FL can be used to train DL models. The MODMA dataset of 3-channel EEG with 26 MDD and 29 non-MDD individuals is used in this paper. Future study on MDD has several potential areas of use and these FL approaches can also be coupled with new technologies. [ABSTRACT FROM AUTHOR]

In view of the inevitable error problems caused by peak extraction, calibration curve fitting and other essential operations in the traditional data processing of chromatic confocal microscopy (CCM), a regression model based on convolutional neural network (CNN) is proposed so that the above necessary operations are no longer required. This CNN-based regression model draws on the core concepts of the AlexNet model and has been moderately customized and optimized to make it more suitable for CCM application scenarios. The proposed method has been v alidated using a completely homemade CCM apparatus The experimental results showed that the CNN-based method is feasible for the CCM measurement and exhibits better stability and higher axial resolution than traditional methods, indicating that deep learning has good application value in the data processing of CCM. • A new chromatic confocal sensing method based on artificial intelligence is proposed. A model based on the deep convolutional neural network (CNN) has been established. • With the new method, traditional data processing methods, i.e., the peak extraction algorithm and the calibration procedure, required in chromatic confocal microscopy are no longer needed. • An experimental prototype has been built, and detailed experimental tests for resolution, accuracy, and 3D topography were carried out. • The measurement results proved that the CNN-based model can improve the axial resolution and measurement precision. [ABSTRACT FROM AUTHOR]

Large-scale crop mapping from dense time-series images is a difficult task and becomes even more challenging with the cloud coverage. Current deep learning models frequently represent time series from a single perspective, which is insufficient to obtain fine-grained details. Meanwhile, the impact of cloud noise on deep learning models is not yet fully understood. In this study, a Multi-scale Temporal Transformer-Conv network (Ms-TTC) is proposed for robust crop mapping under frequently clouds. The Ms-TTC enhances temporal representations by effectively combining the global modeling capability of self-attention with the local capture capability of convolutional neural network (CNN) at multi-temporal scales. The Ms-TTC network consists of three main components: (1) a temporal encoder module that explores global and local temporal relationships at multi-temporal scales, (2) an attention-based fusion module that effectively fuses multi-scale temporal features, and (3) the output module that concatenates the high-level time series features and refined multi-scale features to predict the label. The proposed model demonstrated superior performance compared to state-of-the-art methods on the large-scale time series dataset, FranceCrops, achieving a minimum improvement of 2% in mF1 scores. Subsequently, gradient back-propagation-based feature importance analysis was used to investigate the behavior of deep learning models for processing time series data with cloud noise. The results revealed that most deep learning models can suppress cloudy observations to some degree, and models with a global field of view had superior cloud masking but also lost some local temporal information. Clouds can influence the model's attention towards the spectral dimension, particularly affecting the visible and vegetation red-edge bands, which exhibit higher sensitivity to cloud noise and play a crucial role to performance. This study provides a feasible approach for large-scale dynamic crop mapping independently of cloudy conditions by combining global-local temporal representations at multi-scales. [ABSTRACT FROM AUTHOR]

With diseases and weeds alone accounting for approximately 30% of productivity losses in agriculture worldwide, plant diseases are a significant factor. Farmers frequently consult experts to diagnose their crops in an effort to reduce these losses. On the other hand, conventional visual assessment techniques focused on elements like leaf color, shape, and texture can be inefficient, time-consuming, and expensive. This study suggests using symptom traits like yellow leaves, black lesions, wilting, and plant mortality in crops like apple and grape to anticipate plant diseases, notably black rot. AlexNet, VGGNet, ResNet, and Inception are the four deep learning models examined in the study; VGGNet achieved the greatest accuracy of 98.19%. [ABSTRACT FROM AUTHOR]

Fault diagnosis based on deep learning (DL) has been a research hotspot in recent years. However, the current neural networks are getting larger and larger, with more and more parameters and insufficient noise resistance, making it difficult to effectively apply these methods to real working conditions. To address these issues, we propose a novel deep learning method with fewer parameters and better noise resistance based on transfer adversarial subnetwork (TAS) and channel-wise thresholds (CWT), namely, anti-noise transfer adversarial convolutions (ANTAC). In the proposed method, the original data and feature vectors are mapped to reproducing kernel Hilbert space (RKHS) and processed by maximum mean discrepancy (MMD) and Wasserstein distance (WD), which makes the method more capable to distinguish the similar features without producing any additional training parameters. Furthermore, white Gaussian noise (WGN) and the soft thresholding method with CWT are used to reduce data noise and improve the robustness and noise resistance of the network. Finally, the superiority of the proposed method is validated through experiments on different datasets, network structures and the data with different SNRs. The results show that the proposed method has better feature discrimination ability, noise resistance, and fewer parameters compared with other methods. The highest accuracy of the proposed method is 99.90% on the test set. • In order to enhance the network ability in distinguishing similar types of samples, TAS is proposed. Experiments show that the proposed method can effectively improve the network stability and the feature discrimination ability with fewer parameters. • The anti-interference ability and robustness of the network is enhanced by adding a certain disturbance into it. In addition. The soft threshold function with channel-wise threshold is used to improve the denoising ability of the network. • To enhance the network generalization and sensitivity to similar features, AFV loss is calculated by WD and controlled by scaling factor λ , and finally sent into RMSprop optimizer together with classification loss for optimization. • The method can directly process the original signals and the procedures of denoising and feature extraction are finished automatically in the method, which improves the fault diagnosis efficiency and provides a further practical possibility for the deployment of end-to-end fault diagnosis platforms. [ABSTRACT FROM AUTHOR]