Up-sampling Process Research Articles

A multi-channel spatial information feature based human pose estimation algorithm

Human pose estimation is an important task in computer vision, which can provide key point detection of human body and obtain bone information. At present, human pose estimation is mainly utilized for detection of large targets, and there is no solution for detection of small targets. This paper proposes a multi-channel spatial information feature based human pose (MCSF-Pose) estimation algorithm to address the issue of medium and small targets inaccurate detection of human key points in scenarios involving occlusion and multiple poses. The MCSF-Pose network is a bottom-up regression network. Firstly, an UP-Focus module is designed to expand the feature information while reducing parameter computation during the up-sampling process. Then, the channel segmentation strategy is adopted to cut the features, and the feature information of multiple dimensions is retained through different convolutional groups, which reduces the parameter lightweight network model and makes up for the loss of the feature information associated with the depth of the network. Finally, the three-layer PANet structure is designed to reduce the complexity of the model. With the aid of the structure, it also to improve the detection accuracy and anti-interference ability of human key points. The experimental results indicate that the proposed algorithm outperforms YOLO-Pose and other human pose estimation algorithms on COCO2017 and MPII human pose datasets.

MSRA-Net: multi-channel semantic-aware and residual attention mechanism network for unsupervised 3D image registration

Objective. Convolutional neural network (CNN) is developing rapidly in the field of medical image registration, and the proposed U-Net further improves the precision of registration. However, this method may discard certain important information in the process of encoding and decoding steps, consequently leading to a decline in accuracy. To solve this problem, a multi-channel semantic-aware and residual attention mechanism network (MSRA-Net) is proposed in this paper. Approach. Our proposed network achieves efficient information aggregation by cleverly extracting the features of different channels. Firstly, a context-aware module (CAM) is designed to extract valuable contextual information. And the depth-wise separable convolution is employed in the CAM to alleviate the computational burden. Then, a new multi-channel semantic-aware module (MCSAM) is designed for more comprehensive fusion of up-sampling features. Additionally, the residual attention module is introduced in the up-sampling process to extract more semantic information and minimize information loss. Main results. This study utilizes Dice score, average symmetric surface distance and negative Jacobian determinant evaluation metrics to evaluate the influence of registration. The experimental results demonstrate that our proposed MSRA-Net has the highest accuracy compared to several state-of-the-art methods. Moreover, our network has demonstrated the highest Dice score across multiple datasets, thereby indicating that the superior generalization capabilities of our model. Significance. The proposed MSRA-Net offers a novel approach to improve medical image registration accuracy, with implications for various clinical applications. Our implementation is available at https://github.com/shy922/MSRA-Net.

Image up-sampling method for dual-radiation nondestructive cargo inspection system

In this paper, an image up-sampling method for a dual-radiation nondestructive cargo inspection system, that uses x-ray and neutron, is proposed. The inspection system is assumed to have a larger neutron detector pixel size than that of x-ray detector pixel, which is usually the case to address the lower neutron source flux and the lower detector efficiency than those in the x-ray imaging system. The proposed technique adopts a cluster map of an x-ray projection image with high spatial resolution to up-sample the corresponding neutron projection image with low spatial resolution. The cluster map is acquired by use of a modified k-means clustering algorithm. The up-sampling process is formulated by a linear least-square problem of mapping the neutron projection image onto the cluster map. Consequently, the neutron projection image would have the same dimension with the x-ray projection image array allowing R-value based material decomposition. We have conducted a numerical study using GATE simulation to demonstrate the efficacy of the proposed method. The proposed method was compared with a linear interpolation method and also with an up-sampling technique that does not use the cluster map. The outperformance of the proposed method over those reference techniques has been successfully presented.

An Implicit Transformer-based Fusion Method for Hyperspectral and Multispectral Remote Sensing Image

There is an effective way to enhance the spatial resolution of hyperspectral remote sensing images by fusing them with multispectral remote sensing images. However, most of the existing deep fusion techniques adopt discretized explicit models to approximate the complex continuous nonlinear mapping in the fusion process, leading to limitations in enhancing the fidelity of spatial details. Additionally, existing algorithms commonly utilize discrete methods such as bilinear or bicubic interpolation during the hyperspectral upsampling process, leading to the loss of crucial spatial-spectral features. To this end, this study proposes a novel Implicit Transformer Fusion Generative Adversarial Network (ITF-GAN), which incorporates the continuity perception mechanism of implicit neural representation with the powerful self-attention mechanism of the Transformer architecture, which uses point-to-point implicit functions aiming to efficiently process information in both spatial and spectral dimensions. Besides, a guided implicit neural sampling module is introduced in the hyperspectral image up-sampling process to enhance the coordinated expression of features in the spatial and spectral domains, which improves the spatial resolution and spectral fidelity of the fused image during the upsampling process. A series of fusion experiments including 4x, 8x, and 16x scale factors have shown that ITF-GAN has significant advantages over current popular fusion algorithms in both objective evaluation indicators and subjective visual evaluation.

Global Semantic-Sense Aggregation Network for Salient Object Detection in Remote Sensing Images.

Salient object detection (SOD) aims to accurately identify significant geographical objects in remote sensing images (RSI), providing reliable support and guidance for extensive geographical information analyses and decisions. However, SOD in RSI faces numerous challenges, including shadow interference, inter-class feature confusion, as well as unclear target edge contours. Therefore, we designed an effective Global Semantic-aware Aggregation Network (GSANet) to aggregate salient information in RSI. GSANet computes the information entropy of different regions, prioritizing areas with high information entropy as potential target regions, thereby achieving precise localization and semantic understanding of salient objects in remote sensing imagery. Specifically, we proposed a Semantic Detail Embedding Module (SDEM), which explores the potential connections among multi-level features, adaptively fusing shallow texture details with deep semantic features, efficiently aggregating the information entropy of salient regions, enhancing information content of salient targets. Additionally, we proposed a Semantic Perception Fusion Module (SPFM) to analyze map relationships between contextual information and local details, enhancing the perceptual capability for salient objects while suppressing irrelevant information entropy, thereby addressing the semantic dilution issue of salient objects during the up-sampling process. The experimental results on two publicly available datasets, ORSSD and EORSSD, demonstrated the outstanding performance of our method. The method achieved 93.91% Sα, 98.36% Eξ, and 89.37% Fβ on the EORSSD dataset.

Multi-scale reciprocal consistency learning for semi-supervised left atrial segmentation

Abstract Accurate and enough labels are crucial to constructing an excellent deep-learning framework in processing medical scans. However, the expensive cost of medical labels greatly impedes this process. In this case, semi-supervised learning has shown great potential due to its efficient use of unlabeled data. Therefore, we present a novel multi-scale reciprocal consistency network (RC-NET) to utilize the unlabeled data more efficiently for more accurate 3D left atrial segmentation. Our model consists of a common encoder and two independent decoders with inconsistent up-sampling. The decoders generate feature maps of hidden layers at different resolutions during the up-sampling process. Lower-resolution features typically contain local and detailed information, while higher-resolution features involve global or abstract information. Subsequently, we applied consistency learning to these feature maps. By combining local and global semantic information, the model can obtain a comprehensive understanding of the segmentation targets. The experimental data indicate the MRC-Net outperforms many semi-supervised learning methods. It achieves more accurate segmentation results by efficiently utilizing unlabeled data. This provides a completely new approach to improving semi-supervised learning.

AFMUNet: Attention Feature Fusion Network Based on a U-Shaped Structure for Cloud and Cloud Shadow Detection

Cloud detection technology is crucial in remote sensing image processing. While cloud detection is a mature research field, challenges persist in detecting clouds on reflective surfaces like ice, snow, and sand. Particularly, the detection of cloud shadows remains a significant area of concern within cloud detection technology. To address the above problems, a convolutional self-attention mechanism feature fusion network model based on a U-shaped structure is proposed. The model employs an encoder–decoder structure based on UNet. The encoder performs down-sampling to extract deep features, while the decoder uses up-sampling to reconstruct the feature map. To capture the key features of the image, Channel Spatial Attention Module (CSAM) is introduced in this work. This module incorporates an attention mechanism for adaptive field-of-view adjustments. In the up-sampling process, different channels are selected to obtain rich information. Contextual information is integrated to improve the extraction of edge details. Feature fusion at the same layer between up-sampling and down-sampling is carried out. The Feature Fusion Module (FFM) facilitates the positional distribution of the image on a pixel-by-pixel basis. A clear boundary is distinguished using an innovative loss function. Finally, the experimental results on the dataset GF1_WHU show that the segmentation results of this method are better than the existing methods. Hence, our model is of great significance for practical cloud shadow segmentation.

High-Resolution Remote Sensing Image Change Detection Based on Cross-Mixing Attention Network

With the powerful discriminative capabilities of convolutional neural networks, change detection has achieved significant success. However, current methods either ignore the spatiotemporal dependencies between dual-temporal images or suffer from decreased accuracy due to registration errors. Addressing these challenges, this paper proposes a method for remote sensing image change detection based on the cross-mixing attention network. To minimize the impact of registration errors on change detection results, a feature alignment module (FAM) is specifically developed in this study. The FAM performs spatial transformations on dual-temporal feature maps, achieving the precise spatial alignment of feature pairs and reducing false positive rates in change detection. Additionally, to fully exploit the spatiotemporal relationships between dual-temporal images, a cross-mixing attention module (CMAM) is utilized to extract global channel information, enhancing feature selection capabilities. Furthermore, attentional maps are created to guide the up-sampling process, optimizing feature information. Comprehensive experiments conducted on the LEVIR-CD and SYSU-CD change detection datasets demonstrate that the proposed model achieves F1 scores of 91.06% and 81.88%, respectively, outperforming other comparative models. In conclusion, the proposed model maintains good performance on two datasets and, thus, has good applicability in various change detection tasks.

FDSR: An Interpretable Frequency Division Stepwise Process Based Single-Image Super-Resolution Network.

Deep learning has excelled in single-image super-resolution (SISR) applications, yet the lack of interpretability in most deep learning-based SR networks hinders their applicability, especially in fields like medical imaging that require transparent computation. To address these problems, we present an interpretable frequency division SR network that operates in the image frequency domain. It comprises a frequency division module and a step-wise reconstruction method, which divides the image into different frequencies and performs reconstruction accordingly. We develop a frequency division loss function to ensure that each reconstruction module (ReM) operates solely at one image frequency. These methods establish an interpretable framework for SR networks, visualizing the image reconstruction process and reducing the black box nature of SR networks. Additionally, we revisited the subpixel layer upsampling process by deriving its inverse process and designing a displacement generation module. This interpretable upsampling process incorporates subpixel information and is similar to pre-upsampling frameworks. Furthermore, we develop a new ReM based on interpretable Hessian attention to enhance network performance. Extensive experiments demonstrate that our network, without the frequency division loss, outperforms state-of-the-art methods qualitatively and quantitatively. The inclusion of the frequency division loss enhances the network's interpretability and robustness, and only slightly decreases the PSNR and SSIM metrics by an average of 0.48 dB and 0.0049, respectively.

Artifact noise suppression of particle-field computed tomography based on lightweight residual and enhanced convergence neural network

<sec>The realization of fast and high-quality three-dimensional particle-field image characterization is always highly desired in the areas, such as experimental fluid mechanics and biomedicine, for the micro-particle distribution status in a flow-field can characterize the field properties well. In the particle-field image reconstruction and characterization, a wildly used approach at present is the computed tomography. The great advantage of the computed tomography for particle-field image reconstruction lies in the fact that the full particle spatial distribution can be obtained and presented due to multi-angle sampling.</sec><sec>Recently, with the development and application of deep learning technique in computed tomography, the image quality has been greatly improved by the powerful learning ability of a deep learning network. In addition, the deep learning application also makes it possible to speed up the computed tomographic imaging process from sparse-sampling due to the ability of the network to strongly extract image feature. However, sparse-sampling will lead to insufficient acquirement of the object information during sampling for the computed tomography. Therefore, a sort of artifact noise will emerge and be accompanied with the reconstructed images, and thus severely affecting the image quality. As there is no universal network approach that can be applied to all types of objects in the suppression of artifact noise, it is still a challenge in removing the sparse-sampling-induced artifact noise in the computed tomography now.</sec><sec>Therefore, we propose and develop a specific lightweight residual and enhanced convergence neural network (LREC-net) approach for suppressing the artifact noise in the particle-field computed tomography. In this method, the network input dataset is also optimized in signal-to-noise ratio (SNR) in order to reduce the input noise and ensure the effective particle image feature extraction of the network in the imaging process.</sec><sec>In the design of LREC-net architecture, a five-layer lightweight and dual-residual down-sampling is constructed on the basis of typical U-net and Resnet50, making the LREC-net more suitable for the particle-field image reconstruction. Moreover, a fast feature convergence module for rapid particle-field feature acquirement is added to up-sampling process of the network to further promote the network processing efficiency. Apart from the design of LREC-net network itself, the optimization of network input dataset in SNR of images is achieved by finding a fit image reconstruction algorithm that can produce higher-SNR particle images in the computed tomography. This achievement reduces the input noise as much as possible and ensures effective particle-field feature extraction by the network.</sec><sec>The simulation analysis and experimental test demonstrate the effectiveness of the proposed LREC-net method, which involves the evaluations of SNR changes of the input-output images through the network, the proportion of residual artifact noise as ghost-particles (GPP) in the reconstructed images, and the valid-particle loss proportion (PLP). In contrast to the performances of U-net and Resnet50 under the same imaging conditions, all the data in SNR, GPP and PLP show the great improvement of the image quality due to the application of LREC-net method. Meanwhile, the designed LREC-net method also enhances the network running efficiency to a large extent due to the remarkable reduction of training time. Therefore, this work provides a new and effective approach for developing sparse-sampling-based fast and high-quality particle-field computed tomography.</sec>

PU-Flow: a Point Cloud Upsampling Network with Normalizing Flows.

Point cloud upsampling aims to generate dense point clouds from given sparse ones, which is a challenging task due to the irregular and unordered nature of point sets. To address this issue, we present a novel deep learning-based model, called PU-Flow, which incorporates normalizing flows and weight prediction techniques to produce dense points uniformly distributed on the underlying surface. Specifically, we exploit the invertible characteristics of normalizing flows to transform points between Euclidean and latent spaces and formulate the upsampling process as ensemble of neighbouring points in a latent space, where the ensemble weights are adaptively learned from local geometric context. Extensive experiments show that our method is competitive and, in most test cases, it outperforms state-of-the-art methods in terms of reconstruction quality, proximity-to-surface accuracy, and computation efficiency. The source code will be publicly available at https://github.com/unknownue/puflow.

Uncertainty-driven trustworthy defect detection for high-resolution powder bed images in selective laser melting

Selective laser melting (SLM) is known as one of the most promising metal additive manufacturing technologies, and how to ensure its consistent quality is still a main challenge, which urgently needs to be addressed. The metal powder spreading quality, as the first stage of SLM, has a direct impact on the subsequent forming and it is necessary to be monitored. However, existing deep learning-based methods for monitoring powder spreading quality often suffer from the problem of unreliability. The high-resolution property of powder bed images is often ignored, and the predicted results are not well evaluated. To address the above issues and achieve trustworthy defect detection, this paper proposes an uncertainty-driven trustworthy defect detection method for high-resolution powder bed images in SLM. A super resolution module based on ISDNet is first proposed to refine the upsampling process. Checkpoint ensemble is adopted for better achieving model uncertainty estimation only requiring a single training process, feature reuse helps reduce the computational load, and temperature scaling is used to further calibrate each ensemble member to get a more precise uncertainty. Besides, we design an uncertainty-driven model improvement method to further improve defect detection performance for regions with high uncertainty. Experiments demonstrate the effectiveness of our proposed method, and our work achieves a trustworthy defect detection for high-resolution powder bed images in SLM.

Object Detection for Aircraft Turnover Milestone based on Modified

In the target detection task of aircraft turnaround milestone in foggy scenario, there are some problems such as unstable location of prediction frame boundary, high error detection rate and poor detection effect of small target. A new target detection method BTM-YOLO (Broad-sighted upsample and three-dimensional attention multiple detection head YOLO) is proposed, which is based on YOLOv7 network. Add a small target detection head to improve the ability of small target detection; The up-sampling module OVRAFE is introduced to reduce the information loss in the up-sampling process. Replace CIoU with Median Wise IoU (MWIoU) to suppress the problem of poor sample swelling in data sets. The improved model makes up for the performance shortcomings of small target detection in foggy days, and the average detection accuracy on the real foggy day test set is 76.2%, which is 3.32% higher than that of the original model, basically meeting the task requirements.

Nonlinear Acoustic Holography with Adaptive Sampling.

Accurate and efficient numerical simulation of highly nonlinear ultrasound propagation is essential for a wide range of therapeutic and physical ultrasound applications. However, due to large domain sizes and the generation of higher harmonics, such simulations are computationally challenging, particularly in three-dimensional problems with shock waves. Current numerical methods are based on computationally inefficient uniform meshes that require the highest harmonic to be resolved across the entire domain. To address this challenge, we present an adaptive numerical algorithm for computationally efficient nonlinear acoustic holography. At each propagation step, the algorithm monitors the harmonic content of the acoustic signal and adjusts its discretization parameters accordingly. This enables efficient local resolution of higher harmonics in areas of high nonlinearity while avoiding unnecessary resolution elsewhere. Furthermore, the algorithm actively adapts to the signal's nonlinearity level, eliminating the need for prior reference simulations or information about the spatial distribution of the harmonic content of the acoustic field. The proposed algorithm incorporates an upsampling process in the frequency domain to accommodate the generation of higher harmonics in forward propagation and a downsampling process when higher harmonics are decimated in backward propagation. The efficiency of the algorithm was evaluated for highly nonlinear three-dimensional problems, demonstrating a significant reduction in computational cost with a nearly 50-fold speedup over a uniform mesh implementation. Our findings enable a more rapid and efficient approach to modeling nonlinear high-intensity focused ultrasound wave propagation.

Smoking behavior detection algorithm based on YOLOv8-MNC.

The detection of smoking behavior is an emerging field faced with challenges in identifying small, frequently occluded objects like cigarette butts using existing deep learning technologies. Such challenges have led to unsatisfactory detection accuracy and poor model robustness. To overcome these issues, this paper introduces a novel smoking detection algorithm, YOLOv8-MNC, which builds on the YOLOv8 network and includes a specialized layer for small target detection. The YOLOv8-MNC algorithm employs three key strategies: (1) It utilizes NWD Loss to mitigate the effects of minor deviations in object positions on IoU, thereby enhancing training accuracy; (2) It incorporates the Multi-head Self-Attention Mechanism (MHSA) to bolster the network's global feature learning capacity; and (3) It implements the lightweight general up-sampling operator CARAFE, in place of conventional nearest-neighbor interpolation up-sampling modules, minimizing feature information loss during the up-sampling process. Experimental results from a customized smoking behavior dataset demonstrate significant improvement in detection accuracy. The YOLOv8-MNC model achieved a detection accuracy of 85.887%, signifying a remarkable increase of 5.7% in the mean Average Precision (mAP@0.5) when compared to the previous algorithm. The YOLOv8-MNC algorithm represents a valuable step forward in resolving existing problems in smoking behavior detection. Its enhanced performance in both detection accuracy and robustness indicates potential applicability in related fields, thus illustrating a meaningful advancement in the sphere of smoking behavior detection. Future efforts will focus on refining this technique and exploring its application in broader contexts.

Biological characters identification for hard clam larva based on the improved YOLOX-s

This study presents a real-time identification method for improving the accuracy and efficiency of counting hard clam larvae, overcoming the limitations of traditional manual methods. The proposed approach enhances the YOLOX-S network by incorporating attention mechanism modules into feature layers and up-sampling processes. The proposed model is subjected to comparative evaluations with YOLOv3, YOLOv4, YOLOv5, and Faster R-CNN networks, demonstrating its superior performance. After 150 epochs, the model achieves a training loss of approximately 1.42 and a mean average precision of 96.50%, outperforming models without attention mechanisms. It accurately identifies the quantity, survival rate, length, and width of hard clam larvae, providing valuable insights for quality prediction in the aquaculture industry. These findings contribute to the advancement of shellfish farming practices, offering a precise and efficient tool for larval assessment and promoting sustainable aquaculture development.

SUNet++: A Deep Network with Channel Attention for Small-Scale Object Segmentation on 3D Medical Images

As a deep learning network with an encoder-decoder architecture, UNet and its series of improved versions have been widely used in medical image segmentation with great applications. However, when used to segment targets in 3D medical images such as magnetic resonance imaging (MRI), computed tomography (CT), these models do not model the relevance of images in vertical space, resulting in poor accurate analysis of consecutive slices of the same patient. On the other hand, the large amount of detail lost during the encoding process makes these models incapable of segmenting small-scale tumor targets. Aiming at the scene of small-scale target segmentation in 3D medical images, a fully new neural network model SUNet++ is proposed on the basis of UNet and UNet++. SUNet++ improves the existing models mainly in three aspects: 1) the modeling strategy of slice superposition is used to thoroughly excavate the three dimensional information of the data; 2) by adding an attention mechanism during the decoding process, small scale targets in the picture are retained and amplified; 3) in the up-sampling process, the transposed convolution operation is used to further enhance the effect of the model. In order to verify the effect of the model, we collected and produced a dataset of hyperintensity MRI liver-stage images containing over 400 cases of liver nodules. Experimental results on both public and proprietary datasets demonstrate the superiority of SUNet++ in small-scale target segmentation of three-dimensional medical images.

IAN: Instance-Augmented Net for 3D Instance Segmentation

When aggregating local information from neighbors, prevailing 3D instance segmentation backbones only leverage 3D coordinates to find neighboring points without identifying whether these points are from the same object as the query point, which causes the model to gather excessive noisy features. Besides, traditional backbones fail to fully utilize multi-resolution information. Therefore, previous methods have difficulty in segmenting targets in cluttered scenes. To tackle these issues, we propose Instance-Augmented Net (IAN). The keys to our approach are Instance-Augmented Block (IAB), Instance-Augmented Upsampler (IAU), and Attentive Fusion (AF). In IAB, for each foreground point, we leverage its instance information to filter out noisy neighbors from other objects. We also propose IAU to apply this instance-augmented strategy to the upsampling process. Furthermore, to retain comprehensive information, we upsample multi-resolution feature maps and adopt attention generated by AF to fuse them. Notably, by encoding neighborhood information, AF can generate attention at point-level adaptively. Moreover, to further test the generality of models, we present Clutter and Occlusion (CAO), a new 3D instance segmentation dataset tailored for robotic grasping tasks. Extensive experiments on S3DIS, ScanNet and CAO show the effectiveness of our IAN.

Rice grains and grain impurity segmentation method based on a deep learning algorithm-NAM-EfficientNetv2

An appropriate image segmentation algorithm is required for discriminating between full grains and grain impurities. In this study, a lightweight fully convolutional segmentation algorithm based on NAM EfficientNetV2 was proposed to solve the problem of slow processing speed in mobile terminal equipment owing to limited computing resources and a large number of model parameters and improve the detection accuracy. First, a standardized NAM attention mechanism was introduced to replace the SE attention mechanism used in EfficientNetV2 and the improved NAM-EfficientNetV2 network was used as a feature extraction structure. Then, in the up-sampling process, the multi-scale features output by the shallow network are fused to effectively use low-level semantics to encode spatial details and fully convolutional pixel segmentation technology is used to achieve rice grain and impurity segmentation. Finally, compared with the baseline model, the detection accuracy of the model was further improved on a self-made dataset. The comprehensive evaluation index F1 of rice grain and its impurities were 95.26% and 93.27%, respectively, and the model parameter amount was 20.6 M. Combined with post-processing, detecting an image on the GPU device took an average of 0.103 s and 0.301 s on the CPU device. The experimental results showed that the improved algorithm is more lightweight, which provides a reference for the model to be deployed in mobile terminal equipment to realize the function of real-time detection of grain impurity in the combine harvester.

DSGA-Net: Deeply separable gated transformer and attention strategy for medical image segmentation network

To address the problems of under-segmentation and over-segmentation of small organs in medical image segmentation. We present a novel medical image segmentation network model with Depth Separable Gating Transformer and a Three-branch Attention module (DSGA-Net). Firstly, the model adds a Depth Separable Gated Visual Transformer (DSG-ViT) module into its Encoder to enhance (i) the contextual links among global, local, and channels and (ii) the sensitivity to location information. Secondly, a Mixed Three-branch Attention (MTA) module is proposed to increase the number of features in the up-sampling process. Meanwhile, the loss of feature information is reduced when restoring the feature image to the original image size. By validating Synapse, BraTs2020, and ACDC public datasets, the Dice Similarity Coefficient (DSC) of the results of DSGA-Net reached 81.24%,85.82%, and 91.34%, respectively. Moreover, the Hausdorff Score (HD) decreased to 20.91% and 5.27% on the Synapse and BraTs2020. There are 10.78% and 0.69% decreases compared to the Baseline TransUNet. The experimental results indicate that DSGA-Net achieves better segmentation than most advanced methods.