Convolutional Neural Network Module Research Articles

3D LVCN: A Lightweight Volumetric ConvNet

ABSTRACTIn recent years, with the significant increase in the volume of three‐dimensional medical image data, three‐dimensional medical models have emerged. However, existing methods often require a large number of model parameters to deal with complex medical datasets, leading to high model complexity and significant consumption of computational resources. In order to address these issues, this paper proposes a 3D Lightweight Volume Convolutional Neural Network (3D LVCN), aiming to achieve efficient and accurate volume segmentation. This network architecture combines the design principles of convolutional neural network modules and hierarchical transformers, using large convolutional kernels as the basic framework for feature extraction, while introducing 1 × 1 × 1 convolutional kernels for deep convolution. This improvement not only enhances the computational efficiency of the model but also improves its generalization ability. The pro‐posed model is tested on three challenging public datasets, namely spleen, liver, and lung, from the medical segmentation decathlon. Experimental results show that the proposed model performance has in‐creased from 0.8315 to 0.8673, with a reduction in parameters of approximately 5%. This indicates that compared to currently advanced model structures, our proposed model architecture exhibits significant advantages in segmentation performance.

Cnn-assisted multi-hop graph attention network for hyperspectral image classification

ABSTRACT Recently, the convolutional neural network (CNN) has gained widespread adoption in the hyperspectral image (HSI) classification owing to its remarkable feature extraction capability. However, the fixed acceptance domain of CNN restricts it to Euclidean image data only, making it difficult to capture complex information in hyperspectral data. To overcome this problem, much attention has been paid to the graph attention network (GAT), which can effectively model graph structure and capture complex dependencies between nodes. However, GAT usually acts on superpixel nodes, which may lead to the loss of pixel-level information. To better integrate the advantages of both, we propose a CNN-assisted multi-hop graph attention network (CMGAT) for HSI classification. Specifically, a parallel dual-branch architecture is first constructed to simultaneously capture spectral-spatial features from hyperspectral data at the superpixel and pixel levels using GAT and CNN, respectively. On this basis, the multi-hop and multi-scale mechanisms are further employed to construct a multi-hop GAT module and a multi-scale CNN module to capture diverse feature information. Secondly, an attention module is cascaded before the multi-scale CNN module to improve classification performance. Eventually, the output information from the two branches is weighted and fused to produce the classification result. We performed experiments on four benchmark HSI datasets, including Indian Pines (IP), University of Pavia (UP), Salinas Valley (SV) and WHU-Hi-LongKou (LK). The results demonstrate that the proposed method outperforms several deep learning methods, achieving overall accuracies of 95.67%, 99.04%, 99.55% and 99.51%, respectively, even with fewer training samples.

GFI-YOLOv8: Sika Deer Posture Recognition Target Detection Method Based on YOLOv8.

As the sika deer breeding industry flourishes on a large scale, accurately assessing the health of these animals is of paramount importance. Implementing posture recognition through target detection serves as a vital method for monitoring the well-being of sika deer. This approach allows for a more nuanced understanding of their physical condition, ensuring the industry can maintain high standards of animal welfare and productivity. In order to achieve remote monitoring of sika deer without interfering with the natural behavior of the animals, and to enhance animal welfare, this paper proposes a sika deer individual posture recognition detection algorithm GFI-YOLOv8 based on YOLOv8. Firstly, this paper proposes to add the iAFF iterative attention feature fusion module to the C2f of the backbone network module, replace the original SPPF module with AIFI module, and use the attention mechanism to adjust the feature channel adaptively. This aims to enhance granularity, improve the model's recognition, and enhance understanding of sika deer behavior in complex scenes. Secondly, a novel convolutional neural network module is introduced to improve the efficiency and accuracy of feature extraction, while preserving the model's depth and diversity. In addition, a new attention mechanism module is proposed to expand the receptive field and simplify the model. Furthermore, a new pyramid network and an optimized detection head module are presented to improve the recognition and interpretation of sika deer postures in intricate environments. The experimental results demonstrate that the model achieves 91.6% accuracy in recognizing the posture of sika deer, with a 6% improvement in accuracy and a 4.6% increase in mAP50 compared to YOLOv8n. Compared to other models in the YOLO series, such as YOLOv5n, YOLOv7-tiny, YOLOv8n, YOLOv8s, YOLOv9, and YOLOv10, this model exhibits higher accuracy, and improved mAP50 and mAP50-95 values. The overall performance is commendable, meeting the requirements for accurate and rapid identification of the posture of sika deer. This model proves beneficial for the precise and real-time monitoring of sika deer posture in complex breeding environments and under all-weather conditions.

Transformer-based improved U-net for high-performance underwater polarization imaging

Polarization imaging plays an important role in underwater imaging technology, as it allows for the suppression of backscattering light and thus enhances the quality of underwater images. Furthermore, with advancements in computer hardware technology, deep learning (DL) has experienced rapid development in recent years. However, existing methods based on convolutional neural networks (CNN) have the limitations of local feature extraction and fail to full use of the salient target’s polarization features. In this paper, we propose a transformer-based improved U-Net (TIU-Net) to further improve the performance of underwater polarization imaging. The proposed TIU-Net leverages CNN and transformer module to extract short-range and long-range features respectively for underwater target reconstruction. Meanwhile, by entering multi-dimensional information, each of which has a different emphasis on the expression of target features, to enhance the performance and stability of the reconstruction model. Experimental results on our established polarization underwater dataset show the superiority of our proposed method for underwater imaging, achieving the efficient imaging and high-quality generalization imaging in highly turbid underwater environments.

Prediction of leukemia peptides using convolutional neural network and protein compositions

Leukemia is a type of blood cell cancer that is in the bone marrow’s blood-forming cells. Two types of Leukemia are acute and chronic; acute enhances fast and chronic growth gradually which are further classified into lymphocytic and myeloid leukemias. This work evaluates a unique deep convolutional neural network (CNN) classifier that improves identification precision by carefully examining concatenated peptide patterns. The study uses leukemia protein expression for experiments supporting two different techniques including independence and applied cross-validation. In addition to CNN, multilayer perceptron (MLP), gated recurrent unit (GRU), and recurrent neural network (RNN) are applied. The experimental results show that the CNN model surpasses competitors with its outstanding predictability in independent and cross-validation testing applied on different features extracted from protein expressions such as amino acid composition (AAC) with a group of AAC (GAAC), tripeptide composition (TPC) with a group of TPC (GTPC), and dipeptide composition (DPC) for calculating its accuracies with their receiver operating characteristic (ROC) curve. In independence testing, a feature expression of AAC and a group of GAAC are applied using MLP and CNN modules, and ROC curves are achieved with overall 100% accuracy for the detection of protein patterns. In cross-validation testing, a feature expression on a group of AAC and GAAC patterns achieved 98.33% accuracy which is the highest for the CNN module. Furthermore, ROC curves show a 0.965% extraordinary result for the GRU module. The findings show that the CNN model is excellent at figuring out leukemia illnesses from protein expressions with higher accuracy.

Estimating crop evapotranspiration of wheat-maize rotation system using hybrid convolutional bidirectional Long Short-Term Memory network with grey wolf algorithm in Chinese Loess Plateau region

Accurate estimation of crop evapotranspiration (ET) is essential for the efficient utilization of agricultural water resources, crop production enhancement, and sustainable agricultural development. However, direct measurement of ET is highly expensive, intricate, and time-consuming, highlighting the imperative of establishing a novel model to accurately estimate ET in agricultural ecosystems. To address the above problems, this study proposed a novel model (GWA-CNN-BiLSTM), which incorporates Grey Wolf Algorithm (GWA), Convolutional Neural Network (CNN), and Bidirectional Long Short-Term Memory network (BiLSTM) as a hyperparameter adjuster, feature extractor, and regression component, respectively, to estimate ET built upon various input combinations comprising net solar radiation (Rn), vapor pressure deficit (VPD), average air temperature (Ta), soil water content (SWC), and leaf area index (LAI) about winter wheat-spring maize rotation system during 2012–2020 in the Loess Plateau. Besides, following a comparative assessment within GWA-CNN-BiLSTM, Convolutional Bidirectional Long Short-Term Memory network (CNN-BiLSTM), BiLSTM, Long Short-Term Memory network (LSTM), and Shuttleworth-Wallace (SW) models, the results revealed that GWA-CNN-BiLSTM under varied inputs obtained the superior performance, ranging from 0.562 to 0.957 in determination coefficient (R2), 8.4–41.5 % in relative root mean square error (RRMSE), 0.349 mm d−1 to 1.521 mm d−1 in mean absolute error (MAE), −3.26 % to 14.11 % in percent bias (PBIAS), and 0.820–1.091 in regression coefficient (b0), respectively. Moreover, while the accuracy of BiLSTM over LSTM was evident, its performance was notably improved by the incorporation of the CNN module. Additionally, LSTM-type models under complete input combination present better precision than SW by 29.7−51.4 % in R2, 44.2−76.1 % in RRMSE, and 33.6−63.4 % in MAE, respectively. Furthermore, the accuracy of all models under varied inputs exhibited excellence in winter wheat compared to spring maize, and corresponding improvements ranged 1.4−4.3 % in R2, 5.1−20.1 % in RRMSE, and 3.1−17.9 % in MAE, respectively. Besides, the meteorological factors (Rn, VPD, Ta) proved to be the most important inputs for ET estimation in winter wheat and spring maize. Wherein the importance of SWC exceeded that of LAI in winter wheat, while the opposite trend was observed in spring maize. In brief, GWA-CNN-BiLSTM is the highly recommended model to estimate ET of winter wheat-spring maize rotation system under diverse input data scenarios in the Loess Plateau, which can facilitate to offer valuable assistance in regional agriculture water management decisions.

Radiological image analysis using effective channel extension and fusion network based on COVID CT images

Attention mechanisms demonstrate significant capabilities in improving the performance of COVID CT image classification networks. Existing strategies often employ global pooling to acquire feature vectors, overlooking local feature information. Complex attention modules are then applied to enhance network performance, inevitably leading to increased model complexity. To address these issues, this paper introduces an Effective Channel Expansion and Fusion (ECEF) module specifically designed to enhance the performance of medical COVID CT image classification. The module represents a simple yet effective convolutional neural network module. In the channel direction of the given feature map, the module performs segmentation, conducts global pooling operations separately on the segmented and overall feature maps to obtain corresponding attention vectors, and merges them to produce high-dimensional attention vectors. Channel attention feature vectors are derived along the two pooling operation directions and dynamically fused. The resulting attention vectors are multiplied by the input feature map for adaptive feature refinement. Through these operations, the ECEF module adeptly captures local features in COVID CT images, enhancing the classification performance of medical COVID CT images. As a lightweight and versatile module, ECEF seamlessly integrates into any CNN architecture with negligible overhead. It can undergo end-to-end training alongside basic CNNs. ResNet is selected as the backbone network for empirical studies conducted on COVID CT images, including the COVID-CT dataset. Experimental results indicate that, compared to current advanced channel attention mechanisms, the ECEF module exhibits superior performance in COVID CT image classification tasks. The successful application of the ECEF module establishes a foundation for its widespread use in areas such as COVID CT medical image segmentation and object detection.

Utilizing a CNN-RNN machine learning approach for forecasting time-series outlet fluid temperature monitoring by long-term operation of BHEs system

The Borehole Heat Exchanger (BHE) plays a pivotal role in enhancing heat exchange efficiency within Ground Source Heat Pump (GSHP) systems. The accurate prediction of the BHE's outlet fluid temperature is crucial for optimizing GSHP performance, energy storage, and resource conservation. However, conventional machine learning methods encounter challenges in manual feature extraction, learning complex nonlinear relationships, and adapting to real-world scenarios. To address these limitations, this research proposes a crossbreed model integrating Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN) architectures to forecast long-term outlet fluid temperature in BHE systems. The model framework encompasses data preprocessing, utilizing refined data in the CNN module for temporal feature extraction, subsequently passed to the RNN module to capture sequential and temporal patterns from each dataset. Specifically, the advanced CNN-RNN architecture is designed to establish a comprehensive input-output mapping, leveraging essential input features such as inlet fluid, ambient air, and subsurface temperatures at varying depths (0, 10, and 20 m). Performance evaluation metrics, including R2, RMSE, MAE, and AARE, are employed to compare and assess prediction accuracy across various models, including LSTM, CNN, and SimpleRNN. The obtained results demonstrate the superior performance of the proposed model, achieving an RSME of 0.818, MAE of 0.642, AARE of 0.0305, and an R2 value of 98.75 %. This surpasses the performance of traditional prediction models (LSTM, CNN, and SimpleRNN) by 3.01 %, 5.80 %, and 19.52 %, respectively. Notably, the remarkably low MAE of 0.642 exhibited by a CNN-RNN model underscores its capability to outperform traditional approaches, especially when handling large datasets. These findings emphasize the significance of the developed model in facilitating efficient operation, positioning it as a valuable tool for advancing the long-term sustainability of BHE systems.

HTC-retina: A hybrid retinal diseases classification model using transformer-Convolutional Neural Network from optical coherence tomography images

Retinal diseases are among nowadays major public health issues, deservedly needing advanced computer-aided diagnosis. We propose a hybrid model for multi label classification, whereby seven retinal diseases are automatically classified from Optical Coherence Tomography (OCT) images. We show that, by combining the strengths of Convolutional Neural Networks (CNNs) and Visual Transformers (ViTs), we can produce a more powerful type of model for medical image classification, especially when considering local lesion information such as retinal diseases. CNNs are indeed proved to be efficient at parameter utilization and provide the ability to extract local features and multi-scale feature maps through convolutional operations. On the other hand, ViT’s self-attention procedure allows processing long-range and global dependencies within an image. The paper clearly shows that the hybridization of these complementary capabilities (CNNs-ViTs) presents a high image processing potential that is more robust and efficient. The proposed model adopts a hierarchical CNN module called Convolutional Patch and Token Embedding (CPTE) instead of employing a direct tokenization approach using the raw input OCT image in the transformer. The CPTE module’s role is to incorporate an inductive bias, to reduce the reliance on large-scale datasets, and to address the low-level feature extraction challenges of the ViT. In addition, considering the importance of local lesion information in OCT images, the model relies on a parallel module called Residual Depthwise-Pointwise ConvNet (RDP-ConvNet) for extracting high-level features. RDP-ConvNet utilizes depthwise and pointwise convolution layers within a residual network architecture. The overall performance of the HTC-Retina model was evaluated on three datasets: the OCT-2017, OCT-C8, and OCT-2014 ; outperforming previous established models, achieving accuracy rates of 99.40%, 97.00%, and 99.77%, respectively ; and sensitivity rates of 99.41%, 97.00%, and 99.77%, respectively. Notably, the model showed high performance while maintaining computational efficiency.

Pure large kernel convolutional neural network transformer for medical image registration

Deformable medical image registration is a fundamental and critical task in medical image analysis. Recently, deep learning-based methods have rapidly developed and have shown impressive results in deformable image registration. However, existing approaches still suffer from limitations in registration accuracy or generalization performance. To address these challenges, in this paper, we propose a pure convolutional neural network module (CVTF) to implement hierarchical transformers and enhance the registration performance of medical images. CVTF has a larger convolutional kernel, providing a larger global effective receptive field, which can improve the network’s ability to capture long-range dependencies. In addition, we introduce the spatial interaction attention (SIA) module to compute the interrelationship between the target feature pixel points and all other points in the feature map. This helps to improve the semantic understanding of the model by emphasizing important features and suppressing irrelevant ones. Based on the proposed CVTF and SIA, we construct a novel registration framework named PCTNet. We applied PCTNet to generate displacement fields and register medical images, and we conducted extensive experiments and validation on two public datasets, OASIS and LPBA40. The experimental results demonstrate the effectiveness and generality of our method, showing significant improvements in registration accuracy and generalization performance compared to existing methods. Our code has been available at https://github.com/fz852/PCTNet.

Revolutionizing Deepfake Detection and Realtime Video Vision with CNN-based Deep Learning Model

The rapid advancement of deep learning models that can generate and synthesis hyper-realistic videos known as Deepfakes and their ease of access have raised concern on possible malicious intent use. Deep learning techniques can now generate faces, swap faces between two subjects in a video, alter facial expressions, change gender, and alter facial features, to list a few. These powerful video manipulation methods have potential use in many fields. However, they also pose a looming threat to everyone if used for harmful purposes such as identity theft, phishing, and scam. In this work, we propose a Convolutional Vision Transformer for the detection of Deepfakes. The Convolutional Vision Transformer: Convolutional Neural Network (CNN) and Vision Transformer (ViT). The CNN extracts learnable features while the ViT takes in the learned features as input and categorizes them using an attention mechanism. We trained our model on the DeepFake Detection Challenge Dataset (DFDC) and have achieved 91.5 percent accuracy, an AUC value of 0.91, and a loss value of 0.32. Our contribution is that we have added a CNN module to the ViT architecture and have achieved a competitive result on the DFDC dataset.

IPSI(2L)-EDL: a Two-layer Predictor for Identifying Promoters and their Types based on Ensemble Deep Learning

Abstract: Promoters are DNA fragments located near the transcription initiation site, they can be divided into strong promoter type and weak promoter type according to transcriptional activation and expression level. Identifying promoters and their strengths in DNA sequences is essential for understanding gene expression regulation. Therefore, it is crucial to further improve predictive quality of predictors for real-world application requirements. Here, we constructed the latest training dataset based on the RegalonDB website, where all the promoters in this dataset have been experimentally validated, and their sequence similarity is less than 85%. We used one-hot and nucleotide chemical property and density (NCPD) to represent DNA sequence samples. Additionally, we proposed an ensemble deep learning framework containing a multi-head attention module, long short-term memory present, and a convolutional neural network module. : The results showed that iPSI(2L)-EDL outperformed other existing methods for both promoter prediction and identification of strong promoter type and weak promoter type, the AUC and MCC for the iPSI(2L)-EDL in identifying promoter were improved by 2.23% and 2.96% compared to that of PseDNC-DL on independent testing data, respectively, while the AUC and MCC for the iPSI(2L)- EDL were increased by 3.74% and 5.86% in predicting promoter strength type, respectively. The results of ablation experiments indicate that CNN plays a crucial role in recognizing promoters, the importance of different input positions and long-range dependency relationships among features are helpful for recognizing promoters. : Furthermore, to make it easier for most experimental scientists to get the results they need, a userfriendly web server has been established and can be accessed at http://47.94.248.117/IPSW(2L)-EDL.

UViT-Seg: An Efficient ViT and U-Net-Based Framework for Accurate Colorectal Polyp Segmentation in Colonoscopy and WCE Images.

Colorectal cancer (CRC) stands out as one of the most prevalent global cancers. The accurate localization of colorectal polyps in endoscopy images is pivotal for timely detection and removal, contributing significantly to CRC prevention. The manual analysis of images generated by gastrointestinal screening technologies poses a tedious task for doctors. Therefore, computer vision-assisted cancer detection could serve as an efficient tool for polyp segmentation. Numerous efforts have been dedicated to automating polyp localization, with the majority of studies relying on convolutional neural networks (CNNs) to learn features from polyp images. Despite their success in polyp segmentation tasks, CNNs exhibit significant limitations in precisely determining polyp location and shape due to their sole reliance on learning local features from images. While gastrointestinal images manifest significant variation in their features, encompassing both high- and low-level ones, a framework that combines the ability to learn both features of polyps is desired. This paper introduces UViT-Seg, a framework designed for polyp segmentation in gastrointestinal images. Operating on an encoder-decoder architecture, UViT-Seg employs two distinct feature extraction methods. A vision transformer in the encoder section captures long-range semantic information, while a CNN module, integrating squeeze-excitation and dual attention mechanisms, captures low-level features, focusing on critical image regions. Experimental evaluations conducted on five public datasets, including CVC clinic, ColonDB, Kvasir-SEG, ETIS LaribDB, and Kvasir Capsule-SEG, demonstrate UViT-Seg's effectiveness in polyp localization. To confirm its generalization performance, the model is tested on datasets not used in training. Benchmarking against common segmentation methods and state-of-the-art polyp segmentation approaches, the proposed model yields promising results. For instance, it achieves a mean Dice coefficient of 0.915 and a mean intersection over union of 0.902 on the CVC Colon dataset. Furthermore, UViT-Seg has the advantage of being efficient, requiring fewer computational resources for both training and testing. This feature positions it as an optimal choice for real-world deployment scenarios.

BERT-TFBS: a novel BERT-based model for predicting transcription factor binding sites by transfer learning.

Transcription factors (TFs) are proteins essential for regulating genetic transcriptions by binding to transcription factor binding sites (TFBSs) in DNA sequences. Accurate predictions of TFBSs can contribute to the design and construction of metabolic regulatory systems based on TFs. Although various deep-learning algorithms have been developed for predicting TFBSs, the prediction performance needs to be improved. This paper proposes a bidirectional encoder representations from transformers (BERT)-based model, called BERT-TFBS, to predict TFBSs solely based on DNA sequences. The model consists of a pre-trained BERT module (DNABERT-2), a convolutional neural network (CNN) module, a convolutional block attention module (CBAM) and an output module. The BERT-TFBS model utilizes the pre-trained DNABERT-2 module to acquire the complex long-term dependencies in DNA sequences through a transfer learning approach, and applies the CNN module and the CBAM to extract high-order local features. The proposed model is trained and tested based on 165 ENCODE ChIP-seq datasets. We conducted experiments with model variants, cross-cell-line validations and comparisons with other models. The experimental results demonstrate the effectiveness and generalization capability of BERT-TFBS in predicting TFBSs, and they show that the proposed model outperforms other deep-learning models. The source code for BERT-TFBS is available at https://github.com/ZX1998-12/BERT-TFBS.

Research on lung sound classification model based on dual-channel CNN-LSTM algorithm

ulmonary diseases have a significant impact on human health and life safety, and abnormalities in the lungs are a direct response to lung diseases. Establishing an effective lung sound classification model that can assist in diagnosis is of great significance for electronic auscultation.In addressing the issue of lung sound signal classification, this study introduces a deep learning classification model based on a dual-channel CNN-LSTM algorithm. Initially, Mel-scale Frequency Cepstral Coefficients (MFCC) are employed for feature extraction from the dataset, transforming lung sound signals into Mel spectrograms. On this foundation, a dual-channel algorithm classification model is constructed, with parallel Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) modules. The CNN module is designed to capture spatial dimension features of the input data, while the LSTM module focuses on temporal dimension features. These two feature sets are fused together, enabling the model to classify lung sounds and thereby assisting in diagnosing pulmonary diseases for healthcare practitioners. This experiment used the ICBHI2017 Challenge Lungs dataset and obtained 5054 pieces of data through data augmentation and sampling techniques.The results show that the accuracy, recall, and F1 score of this model reach 99.01%, 99.13%, and 0.9915, respectively, significantly superior to other models, highlighting its practical application value.

CSPFormer: A cross-spatial pyramid transformer for visual place recognition

Recently, the Vision Transformer (ViT), which applied the Transformer structure to various visual detection tasks, has outperformed convolutional neural networks (CNNs). Nonetheless, due to the lack of scale representation ability of the Transformer, how to extract the local features of the scene to effectively form a global descriptor is still a challenging problem. In the paper, we propose a Cross-Spatial Pyramid Transformer (CSPFormer) to learn the discriminative global descriptors from multi-scale visual features for efficient visual place recognition. Specifically, we first develop a pyramid CNN module that can extract multi-scale visual feature representations. Then, the extracted feature representations of multi-scales are input to multiple connected spatial pyramid Transformer modules that adaptively learn the spatial relationship of the different scale descriptors, where the multiple self-attention is applied to learn a global descriptor from discriminative local descriptors. CNN pyramid features and Transformer multi-scale features are mutually weighted to perform cross-spatial feature representation. The multiple self-attention enhances the long-term dependencies of multi-scale visual descriptors and reduces the computational cost. To obtain the final place-matching result accurately, the cosine function is used to calculate the spatial similarity between the two scenes. Experimental results on public place datasets show that the proposed method achieves state-of-the-art on large-scale visual place recognition tasks. Our model has achieved 94.7%, 92.8%, 91.3%, and 95.7% average recall based on the top 1% candidate scenario on KITTI, Nordland, VPRICE, and EuRoc datasets, respectively.

Integrating prior knowledge into a bibranch pyramid network for medical image segmentation

Medical image segmentation is crucial for obtaining accurate diagnoses, and while convolutional neural network (CNN)-based methods have made strides in recent years, they struggle with modeling long-range dependencies. Transformer-based methods improve this task but require more computational resources. The segment anything model (SAM) can generate pixel-level segmentation results for natural images using sparse manual prompts, but it performs poorly on low-contrast, noisy ultrasound images. To address this issue, we propose a new medical image segmentation network architecture that integrates transformer components, CNN modules, and an SAM encoder into a unified framework. This allows us to simultaneously capture both long-range dependencies and local features. Additionally, we incorporate the image features extracted from the SAM model as prior knowledge to achieve further improved segmentation accuracy with limited training data. To reduce the imposed computational stress, we employ an axial attention mechanism to approximate a transformer's effects by expanding the receptive field. Instead of replacing the transformer components with lightweight attention modules, our model is divided into a global branch and a local branch. The global branch extracts context features with the transformer components, while the local branch processes patch tokens with the axial attention mechanism. We also construct an image pyramid to excavate internal statistics and multiscale representations to obtain more accurate segmentation regions. This bibranch pyramid transformer (Bi-BPT) architecture is effective and robust for medical image segmentation, surpassing other related segmentation network architectures. The experimental results obtained on various medical image datasets demonstrate its effectiveness.

Efficient Near-Infrared Spectrum Detection in Nondestructive Wood Testing via Transfer Network Redesign.

This study systematically developed a deep transfer network for near-infrared spectrum detection using convolutional neural network modules as key components. Through meticulous evaluation, specific modules and structures suitable for constructing the near-infrared spectrum detection model were identified, ensuring its effectiveness. This study extensively analyzed the basic network components and explored three unsupervised domain adaptation structures, highlighting their applications in the nondestructive testing of wood. Additionally, five transfer networks were strategically redesigned to substantially enhance their performance. The experimental results showed that the Conditional Domain Adversarial Network and Globalized Loss Optimization Transfer network outperformed the Direct Standardization, Piecewise Direct Standardization, and Spectral Space Transformation models. The coefficients of determination for the Conditional Domain Adversarial Network and Globalized Loss Optimization Transfer network are 82.11% and 83.59%, respectively, with root mean square error prediction values of 12.237 and 11.582, respectively. These achievements represent considerable advancements toward the practical implementation of an efficient and reliable near-infrared spectrum detection system using a deep transfer network.

Approach to Perforation Interval Selection Uses AI-Assisted Image Interpretation

_ This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 216856, “ML-Driven Integrated Approach for Perforation Interval Selection Based on Advanced Borehole-Image AI-Assisted Interpretation,” by Alexander Petrov, SPE, Mounir Belouahchia, SPE, and Abdelwahab Noufal, SPE, ADNOC. The paper has not been peer reviewed. _ In the complete paper, the authors propose an artificial-intelligence (AI)-assisted work flow that uses machine-learning (ML) techniques to identify sweet spots in carbonate reservoirs. This process involves annotation of geologic features using a well database, with supervision from subject-matter experts (SMEs). The resulting ML model is tested on new wells and can identify pay zones, perforation intervals, and stress analysis. The models successfully detect fractures, breakouts, bedding planes, vugs, and slippage passages with pixel-level precision, reducing borehole-image (BHI) analysis time. BHI Interpretation and Preprocessing The use of BHIs requires manual interpretation and data identification, heavily relying on the expertise and time of SMEs. A widely adopted approach to address this challenge is the use of supervised computer-vision algorithms, a subfield of AI. These algorithms optimize the task function or model based on examples they have learned from data during training. However, when applied to BHIs, certain ML challenges must be addressed, including the following: - Detecting features in wells from different reservoirs using a model trained on wells from one reservoir can be highly challenging because reservoirs may exhibit distinct geological characteristics. - The handling of parts of BHIs with missing data, depicted by vertically slanted white strips, poses considerable difficulty. Therefore, the authors created a deep-learning approach based on a generative adversarial network architecture to fill the gaps automatically (Fig. 1). - The labels provided by geologists often do not have pixelwise precision, causing the machine to become confused while trying to learn inconsistent patterns. The authors use a convolutional neural network (CNN) to compute a probability map for pixels belonging to specific classes. In this application, a class is defined as any of the heterogeneities in the BHI; however, this method is applicable to any type of heterogeneities in an image. After training, the CNN module provides the optimal probability for each pixel in the image. To classify regions in the BHI based on heterogeneities, a class-specific threshold is established. Pixels with values above the thresholds are assigned to the corresponding class, while those below the thresholds are assigned to the background. BHI-Derived Porosity A new approach for borehole-derived porosity was developed in-house to overcome the limitations of existing techniques widely used in the industry. This approach capitalizes on BHIs for multiple analyses, including structural dip assessment, fault and fracture identification, and determination of minimum and maximum horizontal stress orientation. However, its primary strength lies in quantifying the fraction of secondary porosity in heterogeneous or dual-porosity carbonate formations. The authors have devised a novel method that uses borehole electrical images to conduct porosity and image connectivity analysis. By implementing this technique, essential information can be extracted regarding the spatial distribution of porosity and the extent of secondary porosity fraction.

Design of 3D Environment Combining Digital Image Processing Technology and Convolutional Neural Network

As virtual reality technology advances, 3D environment design and modeling have garnered increasing attention. Applications in networked virtual environments span urban planning, industrial design, and manufacturing, among other fields. However, existing 3D modeling methods exhibit high reconstruction error precision, limiting their practicality in many domains, particularly environmental design. To enhance 3D reconstruction accuracy, this study proposes a digital image processing technology that combines binocular camera calibration, stereo correction, and a convolutional neural network (CNN) algorithm for optimization and improvement. By employing the refined stereo-matching algorithm, a 3D reconstruction model was developed to augment 3D environment design and reconstruction accuracy while optimizing the 3D reconstruction effect. An experiment using the ShapeNet dataset demonstrated that the evaluation indices—Chamfer distance (CD), Earth mover’s distance (EMD), and intersection over union—of the model constructed in this study outperformed those of alternative methods. After incorporating the CNN module in the ablation experiment, CD and EMD increased by an average of 0.1 and 0.06, respectively. This validates that the proposed CNN module effectively enhances point cloud reconstruction accuracy. Upon adding the CNN module, the CD index and EMD index in the dataset increased by an average of 0.34 and 0.54, respectively. These results indicate that the proposed CNN module exhibits strong predictive capabilities for point cloud coordinates. Furthermore, the model demonstrates good generalization performance.