The application of transformer networks and feature fusion models in medical image segmentation has aroused considerable attention within the academic circle.Nevertheless,two main obstacles persist:(1)the restrictions...The application of transformer networks and feature fusion models in medical image segmentation has aroused considerable attention within the academic circle.Nevertheless,two main obstacles persist:(1)the restrictions of the Transformer network in dealing with locally detailed features,and(2)the considerable loss of feature information in current feature fusion modules.To solve these issues,this study initially presents a refined feature extraction approach,employing a double-branch feature extraction network to capture complex multi-scale local and global information from images.Subsequently,we proposed a low-loss feature fusion method-Multi-branch Feature Fusion Enhancement Module(MFFEM),which realizes effective feature fusion with minimal loss.Simultaneously,the cross-layer cross-attention fusion module(CLCA)is adopted to further achieve adequate feature fusion by enhancing the interaction between encoders and decoders of various scales.Finally,the feasibility of our method was verified using the Synapse and ACDC datasets,demonstrating its competitiveness.The average DSC(%)was 83.62 and 91.99 respectively,and the average HD95(mm)was reduced to 19.55 and 1.15 respectively.展开更多
Sputum smear tests are critical for the diagnosis of respiratory diseases. Automatic segmentation of bacteria from spu-tum smear images is important for improving diagnostic efficiency. However, this remains a challen...Sputum smear tests are critical for the diagnosis of respiratory diseases. Automatic segmentation of bacteria from spu-tum smear images is important for improving diagnostic efficiency. However, this remains a challenging task owing to the high interclass similarity among different categories of bacteria and the low contrast of the bacterial edges. To explore more levels of global pattern features to promote the distinguishing ability of bacterial categories and main-tain sufficient local fine-grained features to ensure accurate localization of ambiguous bacteria simultaneously, we propose a novel dual-branch deformable cross-attention fusion network (DB-DCAFN) for accurate bacterial segmen-tation. Specifically, we first designed a dual-branch encoder consisting of multiple convolution and transformer blocks in parallel to simultaneously extract multilevel local and global features. We then designed a sparse and deformable cross-attention module to capture the semantic dependencies between local and global features, which can bridge the semantic gap and fuse features effectively. Furthermore, we designed a feature assignment fusion module to enhance meaningful features using an adaptive feature weighting strategy to obtain more accurate segmentation. We conducted extensive experiments to evaluate the effectiveness of DB-DCAFN on a clinical dataset comprising three bacterial categories: Acinetobacter baumannii, Klebsiella pneumoniae, and Pseudomonas aeruginosa. The experi-mental results demonstrate that the proposed DB-DCAFN outperforms other state-of-the-art methods and is effective at segmenting bacteria from sputum smear images.展开更多
In the realm of low-level vision tasks,such as image deraining and dehazing,restoring images distorted by adverse weather conditions remains a significant challenge.The emergence of abundant computational resources ha...In the realm of low-level vision tasks,such as image deraining and dehazing,restoring images distorted by adverse weather conditions remains a significant challenge.The emergence of abundant computational resources has driven the dominance of deep Convolutional Neural Networks(CNNs),supplanting traditional methods reliant on prior knowledge.However,the evolution of CNN architectures has tended towards increasing complexity,utilizing intricate structures to enhance performance,often at the expense of computational efficiency.In response,we propose the Selective Kernel Dense Residual M-shaped Network(SKDRMNet),a flexible solution adept at balancing computational efficiency with network accuracy.A key innovation is the incorporation of an M-shaped hierarchical structure,derived from the U-Net framework as M-Network(M-Net),within which the Selective Kernel Dense Residual Module(SDRM)is introduced to reinforce multi-scale semantic feature maps.Our methodology employs two sampling techniques-bilinear and pixel unshuffled and utilizes a multi-scale feature fusion approach to distil more robust spatial feature map information.During the reconstruction phase,feature maps of varying resolutions are seamlessly integrated,and the extracted features are effectively merged using the Selective Kernel Fusion Module(SKFM).Empirical results demonstrate the comprehensive superiority of SKDRMNet across both synthetic and real rain and haze datasets.展开更多
With the development of smart agriculture,accurately identifying crop diseases through visual recognition techniques instead of by eye has been a significant challenge.This study focused on apple leaf disease,which is...With the development of smart agriculture,accurately identifying crop diseases through visual recognition techniques instead of by eye has been a significant challenge.This study focused on apple leaf disease,which is closely related to the final yield of apples.A multiscale fusion dense network combined with an efficient multiscale attention(EMA)mechanism called Incept_EMA_DenseNet was developed to better identify eight complex apple leaf disease images.Incept_EMA_DenseNet consists of three crucial parts:the inception module,which substituted the convolution layer with multiscale fusion methods in the shallow feature extraction layer;the EMA mechanism,which is used for obtaining appropriate weights of different dense blocks;and the improved DenseNet based on DenseNet_121.Specifically,to find appropriate multiscale fusion methods,the residual module and inception module were compared to determine the performance of each technique,and Incept_EMA_DenseNet achieved an accuracy of 95.38%.Second,this work used three attention mechanisms,and the efficient multiscale attention mechanism obtained the best performance.Third,the convolution layers and bottlenecks were modified without performance degradation,reducing half of the computational load compared with the original models.Incept_EMA_DenseNet,as proposed in this paper,has an accuracy of 96.76%,being 2.93%,3.44%,and 4.16%better than Resnet50,DenseNet_121 and GoogLeNet,respectively,proved to be reliable and beneficial,and can effectively and conveniently assist apple growers with leaf disease identification in the field.展开更多
Time series data plays a crucial role in intelligent transportation systems.Traffic flow forecasting represents a precise estimation of future traffic flow within a specific region and time interval.Existing approache...Time series data plays a crucial role in intelligent transportation systems.Traffic flow forecasting represents a precise estimation of future traffic flow within a specific region and time interval.Existing approaches,including sequence periodic,regression,and deep learning models,have shown promising results in short-term series forecasting.However,forecasting scenarios specifically focused on holiday traffic flow present unique challenges,such as distinct traffic patterns during vacations and the increased demand for long-term forecastings.Consequently,the effectiveness of existing methods diminishes in such scenarios.Therefore,we propose a novel longterm forecasting model based on scene matching and embedding fusion representation to forecast long-term holiday traffic flow.Our model comprises three components:the similar scene matching module,responsible for extracting Similar Scene Features;the long-short term representation fusion module,which integrates scenario embeddings;and a simple fully connected layer at the head for making the final forecasting.Experimental results on real datasets demonstrate that our model outperforms other methods,particularly in medium and long-term forecasting scenarios.展开更多
The Visual tracking problem can usually be solved in two parts.The first part is to extract the feature of the target and get the candidate region.The second part is to realize the classification of the target and the...The Visual tracking problem can usually be solved in two parts.The first part is to extract the feature of the target and get the candidate region.The second part is to realize the classification of the target and the regression of the bounding box.In recent years,Siameses network in visual tracking problem has always been a frontier research hotspot.In this work,it applies two branches namely search area and tracking template area for similar learning to track.Some related researches prove the feasibility of this network structure.According to the characteristics of two branch shared networks in Siamese network,we also propos a new fully convolutional Siamese network to solve the visual tracking problem.Based on the Siamese network structure,the network we designed adopt a new fusion module,which realizes the fusion of multiple feature layers at different depths.We also devise a better target state estimation criterion.The overall structure is simple,efficient and has wide applicability.We extensive experiments on challenging benchmarks including generic object tracking-10k(GOT-10K),online object tracking benckmark2015(OTB2015)and unmanned air vehicle123(UAV123),and comparisons with state-of-the-art trackers and the fusion module commonly used in the past,Finally,our network performed better under the same backbone,and achieved good tracking effect,which proved the effectiveness and universality of our designed network and feature fusion method.展开更多
The numerous photos captured by low-price Internet of Things(IoT)sensors are frequently affected by meteorological factors,especially rainfall.It causes varying sizes of white streaks on the image,destroying the image...The numerous photos captured by low-price Internet of Things(IoT)sensors are frequently affected by meteorological factors,especially rainfall.It causes varying sizes of white streaks on the image,destroying the image texture and ruining the performance of the outdoor computer vision system.Existing methods utilise training with pairs of images,which is difficult to cover all scenes and leads to domain gaps.In addition,the network structures adopt deep learning to map rain images to rain-free images,failing to use prior knowledge effectively.To solve these problems,we introduce a single image derain model in edge computing that combines prior knowledge of rain patterns with the learning capability of the neural network.Specifically,the algorithm first uses Residue Channel Prior to filter out the rainfall textural features then it uses the Feature Fusion Module to fuse the original image with the background feature information.This results in a pre-processed image which is fed into Half Instance Net(HINet)to recover a high-quality rain-free image with a clear and accurate structure,and the model does not rely on any rainfall assumptions.Experimental results on synthetic and real-world datasets show that the average peak signal-to-noise ratio of the model decreases by 0.37 dB on the synthetic dataset and increases by 0.43 dB on the real-world dataset,demonstrating that a combined model reduces the gap between synthetic data and natural rain scenes,improves the generalization ability of the derain network,and alleviates the overfitting problem.展开更多
The Indian Test Blanket Module（TBM） program in ITER is one of the major steps in its fusion reactor program towards DEMO and the future fusion power reactor vision. Research and development（RD） is focused on two t...The Indian Test Blanket Module（TBM） program in ITER is one of the major steps in its fusion reactor program towards DEMO and the future fusion power reactor vision. Research and development（RD） is focused on two types of breeding blanket concepts： lead–lithium ceramic breeder（LLCB） and helium-cooled ceramic breeder（HCCB） blanket systems for the DEMO reactor. As part of the ITER-TBM program, the LLCB concept will be tested in one-half of ITER port no. 2, whose materials and technologies will be tested during ITER operation. The HCCB concept is a variant of the solid breeder blanket, which is presently part of our domestic RD program for DEMO relevant technology development. In the HCCB concept Li_2TiO_3 and beryllium are used as the tritium breeder and neutron multiplier, respectively, in the form of a packed bed having edge-on configuration with reduced activation ferritic martensitic steel as the structural material. In this paper two design schemes, mainly two different orientations of pebble beds, are discussed. In the current concept（case-1）, the ceramic breeder beds are kept horizontal in the toroidal–radial direction. Due to gravity, the pebbles may settle down at the bottom and create a finite gap between the pebbles and the top cooling plate, which will affect the heat transfer between them. In the alternate design concept（case-2）, the pebble bed is vertically（poloidal–radial） orientated where the side plates act as cooling plates instead of top and bottom plates. These two design variants are analyzed analytically and 2 D thermal-hydraulic simulation studies are carried out with ANSYS, using the heat loads obtained from neutronic calculations.Based on the analysis the performance is compared and details of the thermal and radiative heat transfer studies are also discussed in this paper.展开更多
Precise polyp segmentation is vital for the early diagnosis and prevention of colorectal cancer(CRC)in clinical practice.However,due to scale variation and blurry polyp boundaries,it is still a challenging task to ach...Precise polyp segmentation is vital for the early diagnosis and prevention of colorectal cancer(CRC)in clinical practice.However,due to scale variation and blurry polyp boundaries,it is still a challenging task to achieve satisfactory segmentation performance with different scales and shapes.In this study,we present a novel edge-aware feature aggregation network(EFA-Net)for polyp segmentation,which can fully make use of cross-level and multi-scale features to enhance the performance of polyp segmentation.Specifically,we first present an edge-aware guidance module(EGM)to combine the low-level features with the high-level features to learn an edge-enhanced feature,which is incorporated into each decoder unit using a layer-by-layer strategy.Besides,a scale-aware convolution module(SCM)is proposed to learn scale-aware features by using dilated convolutions with different ratios,in order to effectively deal with scale variation.Further,a cross-level fusion module(CFM)is proposed to effectively integrate the cross-level features,which can exploit the local and global contextual information.Finally,the outputs of CFMs are adaptively weighted by using the learned edge-aware feature,which are then used to produce multiple side-out segmentation maps.Experimental results on five widely adopted colonoscopy datasets show that our EFA-Net outperforms state-of-the-art polyp segmentation methods in terms of generalization and effectiveness.Our implementation code and segmentation maps will be publicly at https://github.com/taozh2017/EFANet.展开更多
In order to meet the requirements of accurate identification of surface defects on copper strip in industrial production,a detection model of surface defects based on machine vision,CSC-YOLO,is proposed.The model uses...In order to meet the requirements of accurate identification of surface defects on copper strip in industrial production,a detection model of surface defects based on machine vision,CSC-YOLO,is proposed.The model uses YOLOv4-tiny as the benchmark network.First,K-means clustering is introduced into the benchmark network to obtain anchor frames that match the self-built dataset.Second,a cross-region fusion module is introduced in the backbone network to solve the difficult target recognition problem by fusing contextual semantic information.Third,the spatial pyramid pooling-efficient channel attention network(SPP-E)module is introduced in the path aggregation network(PANet)to enhance the extraction of features.Fourth,to prevent the loss of channel information,a lightweight attention mechanism is introduced to improve the performance of the network.Finally,the performance of the model is improved by adding adjustment factors to correct the loss function for the dimensional characteristics of the surface defects.CSC-YOLO was tested on the self-built dataset of surface defects in copper strip,and the experimental results showed that the mAP of the model can reach 93.58%,which is a 3.37% improvement compared with the benchmark network,and FPS,although decreasing compared with the benchmark network,reached 104.CSC-YOLO takes into account the real-time requirements of copper strip production.The comparison experiments with Faster RCNN,SSD300,YOLOv3,YOLOv4,Resnet50-YOLOv4,YOLOv5s,YOLOv7,and other algorithms show that the algorithm obtains a faster computation speed while maintaining a higher detection accuracy.展开更多
基金funded by the Henan Science and Technology research project(222103810042)Support by the open project of scientific research platform of grain information processing center of Henan University of Technology(KFJJ-2021-108)+1 种基金Support by the innovative funds plan of Henan University of Technology(2021ZKCJ14)Henan University of Technology Youth Backbone Teacher Program.
文摘The application of transformer networks and feature fusion models in medical image segmentation has aroused considerable attention within the academic circle.Nevertheless,two main obstacles persist:(1)the restrictions of the Transformer network in dealing with locally detailed features,and(2)the considerable loss of feature information in current feature fusion modules.To solve these issues,this study initially presents a refined feature extraction approach,employing a double-branch feature extraction network to capture complex multi-scale local and global information from images.Subsequently,we proposed a low-loss feature fusion method-Multi-branch Feature Fusion Enhancement Module(MFFEM),which realizes effective feature fusion with minimal loss.Simultaneously,the cross-layer cross-attention fusion module(CLCA)is adopted to further achieve adequate feature fusion by enhancing the interaction between encoders and decoders of various scales.Finally,the feasibility of our method was verified using the Synapse and ACDC datasets,demonstrating its competitiveness.The average DSC(%)was 83.62 and 91.99 respectively,and the average HD95(mm)was reduced to 19.55 and 1.15 respectively.
基金the Natural Science Foundation of Shandong Province,No.ZR2021MH213and in part by the Suzhou Science and Technology Bureau,No.SJC2021023.
文摘Sputum smear tests are critical for the diagnosis of respiratory diseases. Automatic segmentation of bacteria from spu-tum smear images is important for improving diagnostic efficiency. However, this remains a challenging task owing to the high interclass similarity among different categories of bacteria and the low contrast of the bacterial edges. To explore more levels of global pattern features to promote the distinguishing ability of bacterial categories and main-tain sufficient local fine-grained features to ensure accurate localization of ambiguous bacteria simultaneously, we propose a novel dual-branch deformable cross-attention fusion network (DB-DCAFN) for accurate bacterial segmen-tation. Specifically, we first designed a dual-branch encoder consisting of multiple convolution and transformer blocks in parallel to simultaneously extract multilevel local and global features. We then designed a sparse and deformable cross-attention module to capture the semantic dependencies between local and global features, which can bridge the semantic gap and fuse features effectively. Furthermore, we designed a feature assignment fusion module to enhance meaningful features using an adaptive feature weighting strategy to obtain more accurate segmentation. We conducted extensive experiments to evaluate the effectiveness of DB-DCAFN on a clinical dataset comprising three bacterial categories: Acinetobacter baumannii, Klebsiella pneumoniae, and Pseudomonas aeruginosa. The experi-mental results demonstrate that the proposed DB-DCAFN outperforms other state-of-the-art methods and is effective at segmenting bacteria from sputum smear images.
文摘In the realm of low-level vision tasks,such as image deraining and dehazing,restoring images distorted by adverse weather conditions remains a significant challenge.The emergence of abundant computational resources has driven the dominance of deep Convolutional Neural Networks(CNNs),supplanting traditional methods reliant on prior knowledge.However,the evolution of CNN architectures has tended towards increasing complexity,utilizing intricate structures to enhance performance,often at the expense of computational efficiency.In response,we propose the Selective Kernel Dense Residual M-shaped Network(SKDRMNet),a flexible solution adept at balancing computational efficiency with network accuracy.A key innovation is the incorporation of an M-shaped hierarchical structure,derived from the U-Net framework as M-Network(M-Net),within which the Selective Kernel Dense Residual Module(SDRM)is introduced to reinforce multi-scale semantic feature maps.Our methodology employs two sampling techniques-bilinear and pixel unshuffled and utilizes a multi-scale feature fusion approach to distil more robust spatial feature map information.During the reconstruction phase,feature maps of varying resolutions are seamlessly integrated,and the extracted features are effectively merged using the Selective Kernel Fusion Module(SKFM).Empirical results demonstrate the comprehensive superiority of SKDRMNet across both synthetic and real rain and haze datasets.
基金fully supported by the National Natural Science Foundation of China(52072412)。
文摘With the development of smart agriculture,accurately identifying crop diseases through visual recognition techniques instead of by eye has been a significant challenge.This study focused on apple leaf disease,which is closely related to the final yield of apples.A multiscale fusion dense network combined with an efficient multiscale attention(EMA)mechanism called Incept_EMA_DenseNet was developed to better identify eight complex apple leaf disease images.Incept_EMA_DenseNet consists of three crucial parts:the inception module,which substituted the convolution layer with multiscale fusion methods in the shallow feature extraction layer;the EMA mechanism,which is used for obtaining appropriate weights of different dense blocks;and the improved DenseNet based on DenseNet_121.Specifically,to find appropriate multiscale fusion methods,the residual module and inception module were compared to determine the performance of each technique,and Incept_EMA_DenseNet achieved an accuracy of 95.38%.Second,this work used three attention mechanisms,and the efficient multiscale attention mechanism obtained the best performance.Third,the convolution layers and bottlenecks were modified without performance degradation,reducing half of the computational load compared with the original models.Incept_EMA_DenseNet,as proposed in this paper,has an accuracy of 96.76%,being 2.93%,3.44%,and 4.16%better than Resnet50,DenseNet_121 and GoogLeNet,respectively,proved to be reliable and beneficial,and can effectively and conveniently assist apple growers with leaf disease identification in the field.
基金funded by the Natural Science Foundation of Zhejiang Province of China under Grant (No.LY21F020003)Zhejiang Science and Technology Plan Project (No.2021C02060)the Scientific Research Foundation of Hangzhou City University (No.X-202206).
文摘Time series data plays a crucial role in intelligent transportation systems.Traffic flow forecasting represents a precise estimation of future traffic flow within a specific region and time interval.Existing approaches,including sequence periodic,regression,and deep learning models,have shown promising results in short-term series forecasting.However,forecasting scenarios specifically focused on holiday traffic flow present unique challenges,such as distinct traffic patterns during vacations and the increased demand for long-term forecastings.Consequently,the effectiveness of existing methods diminishes in such scenarios.Therefore,we propose a novel longterm forecasting model based on scene matching and embedding fusion representation to forecast long-term holiday traffic flow.Our model comprises three components:the similar scene matching module,responsible for extracting Similar Scene Features;the long-short term representation fusion module,which integrates scenario embeddings;and a simple fully connected layer at the head for making the final forecasting.Experimental results on real datasets demonstrate that our model outperforms other methods,particularly in medium and long-term forecasting scenarios.
文摘The Visual tracking problem can usually be solved in two parts.The first part is to extract the feature of the target and get the candidate region.The second part is to realize the classification of the target and the regression of the bounding box.In recent years,Siameses network in visual tracking problem has always been a frontier research hotspot.In this work,it applies two branches namely search area and tracking template area for similar learning to track.Some related researches prove the feasibility of this network structure.According to the characteristics of two branch shared networks in Siamese network,we also propos a new fully convolutional Siamese network to solve the visual tracking problem.Based on the Siamese network structure,the network we designed adopt a new fusion module,which realizes the fusion of multiple feature layers at different depths.We also devise a better target state estimation criterion.The overall structure is simple,efficient and has wide applicability.We extensive experiments on challenging benchmarks including generic object tracking-10k(GOT-10K),online object tracking benckmark2015(OTB2015)and unmanned air vehicle123(UAV123),and comparisons with state-of-the-art trackers and the fusion module commonly used in the past,Finally,our network performed better under the same backbone,and achieved good tracking effect,which proved the effectiveness and universality of our designed network and feature fusion method.
基金supported by the National Natural Science Foundation of China under Grant no.41975183,and Grant no.41875184 and Supported by a grant from State Key Laboratory of Resources and Environmental Information System.
文摘The numerous photos captured by low-price Internet of Things(IoT)sensors are frequently affected by meteorological factors,especially rainfall.It causes varying sizes of white streaks on the image,destroying the image texture and ruining the performance of the outdoor computer vision system.Existing methods utilise training with pairs of images,which is difficult to cover all scenes and leads to domain gaps.In addition,the network structures adopt deep learning to map rain images to rain-free images,failing to use prior knowledge effectively.To solve these problems,we introduce a single image derain model in edge computing that combines prior knowledge of rain patterns with the learning capability of the neural network.Specifically,the algorithm first uses Residue Channel Prior to filter out the rainfall textural features then it uses the Feature Fusion Module to fuse the original image with the background feature information.This results in a pre-processed image which is fed into Half Instance Net(HINet)to recover a high-quality rain-free image with a clear and accurate structure,and the model does not rely on any rainfall assumptions.Experimental results on synthetic and real-world datasets show that the average peak signal-to-noise ratio of the model decreases by 0.37 dB on the synthetic dataset and increases by 0.43 dB on the real-world dataset,demonstrating that a combined model reduces the gap between synthetic data and natural rain scenes,improves the generalization ability of the derain network,and alleviates the overfitting problem.
文摘The Indian Test Blanket Module（TBM） program in ITER is one of the major steps in its fusion reactor program towards DEMO and the future fusion power reactor vision. Research and development（RD） is focused on two types of breeding blanket concepts： lead–lithium ceramic breeder（LLCB） and helium-cooled ceramic breeder（HCCB） blanket systems for the DEMO reactor. As part of the ITER-TBM program, the LLCB concept will be tested in one-half of ITER port no. 2, whose materials and technologies will be tested during ITER operation. The HCCB concept is a variant of the solid breeder blanket, which is presently part of our domestic RD program for DEMO relevant technology development. In the HCCB concept Li_2TiO_3 and beryllium are used as the tritium breeder and neutron multiplier, respectively, in the form of a packed bed having edge-on configuration with reduced activation ferritic martensitic steel as the structural material. In this paper two design schemes, mainly two different orientations of pebble beds, are discussed. In the current concept（case-1）, the ceramic breeder beds are kept horizontal in the toroidal–radial direction. Due to gravity, the pebbles may settle down at the bottom and create a finite gap between the pebbles and the top cooling plate, which will affect the heat transfer between them. In the alternate design concept（case-2）, the pebble bed is vertically（poloidal–radial） orientated where the side plates act as cooling plates instead of top and bottom plates. These two design variants are analyzed analytically and 2 D thermal-hydraulic simulation studies are carried out with ANSYS, using the heat loads obtained from neutronic calculations.Based on the analysis the performance is compared and details of the thermal and radiative heat transfer studies are also discussed in this paper.
基金supported in part by National Natural Science Foundation of China(Nos.62172228,62201263,62106043 and 62201265).
文摘Precise polyp segmentation is vital for the early diagnosis and prevention of colorectal cancer(CRC)in clinical practice.However,due to scale variation and blurry polyp boundaries,it is still a challenging task to achieve satisfactory segmentation performance with different scales and shapes.In this study,we present a novel edge-aware feature aggregation network(EFA-Net)for polyp segmentation,which can fully make use of cross-level and multi-scale features to enhance the performance of polyp segmentation.Specifically,we first present an edge-aware guidance module(EGM)to combine the low-level features with the high-level features to learn an edge-enhanced feature,which is incorporated into each decoder unit using a layer-by-layer strategy.Besides,a scale-aware convolution module(SCM)is proposed to learn scale-aware features by using dilated convolutions with different ratios,in order to effectively deal with scale variation.Further,a cross-level fusion module(CFM)is proposed to effectively integrate the cross-level features,which can exploit the local and global contextual information.Finally,the outputs of CFMs are adaptively weighted by using the learned edge-aware feature,which are then used to produce multiple side-out segmentation maps.Experimental results on five widely adopted colonoscopy datasets show that our EFA-Net outperforms state-of-the-art polyp segmentation methods in terms of generalization and effectiveness.Our implementation code and segmentation maps will be publicly at https://github.com/taozh2017/EFANet.
基金the Key Project of Basic Research of Yunnan Province(No.202101AS070016)。
文摘In order to meet the requirements of accurate identification of surface defects on copper strip in industrial production,a detection model of surface defects based on machine vision,CSC-YOLO,is proposed.The model uses YOLOv4-tiny as the benchmark network.First,K-means clustering is introduced into the benchmark network to obtain anchor frames that match the self-built dataset.Second,a cross-region fusion module is introduced in the backbone network to solve the difficult target recognition problem by fusing contextual semantic information.Third,the spatial pyramid pooling-efficient channel attention network(SPP-E)module is introduced in the path aggregation network(PANet)to enhance the extraction of features.Fourth,to prevent the loss of channel information,a lightweight attention mechanism is introduced to improve the performance of the network.Finally,the performance of the model is improved by adding adjustment factors to correct the loss function for the dimensional characteristics of the surface defects.CSC-YOLO was tested on the self-built dataset of surface defects in copper strip,and the experimental results showed that the mAP of the model can reach 93.58%,which is a 3.37% improvement compared with the benchmark network,and FPS,although decreasing compared with the benchmark network,reached 104.CSC-YOLO takes into account the real-time requirements of copper strip production.The comparison experiments with Faster RCNN,SSD300,YOLOv3,YOLOv4,Resnet50-YOLOv4,YOLOv5s,YOLOv7,and other algorithms show that the algorithm obtains a faster computation speed while maintaining a higher detection accuracy.
