At present, salient object detection (SOD) has achieved considerable progress. However, the methods that perform well still face the issue of inadequate detection accuracy. For example, sometimes there are problems of...At present, salient object detection (SOD) has achieved considerable progress. However, the methods that perform well still face the issue of inadequate detection accuracy. For example, sometimes there are problems of missed and false detections. Effectively optimizing features to capture key information and better integrating different levels of features to enhance their complementarity are two significant challenges in the domain of SOD. In response to these challenges, this study proposes a novel SOD method based on multi-strategy feature optimization. We propose the multi-size feature extraction module (MSFEM), which uses the attention mechanism, the multi-level feature fusion, and the residual block to obtain finer features. This module provides robust support for the subsequent accurate detection of the salient object. In addition, we use two rounds of feature fusion and the feedback mechanism to optimize the features obtained by the MSFEM to improve detection accuracy. The first round of feature fusion is applied to integrate the features extracted by the MSFEM to obtain more refined features. Subsequently, the feedback mechanism and the second round of feature fusion are applied to refine the features, thereby providing a stronger foundation for accurately detecting salient objects. To improve the fusion effect, we propose the feature enhancement module (FEM) and the feature optimization module (FOM). The FEM integrates the upper and lower features with the optimized features obtained by the FOM to enhance feature complementarity. The FOM uses different receptive fields, the attention mechanism, and the residual block to more effectively capture key information. Experimental results demonstrate that our method outperforms 10 state-of-the-art SOD methods.展开更多
The Quadric Error Metrics(QEM)algorithm is a widely used method for mesh simplification;however,it often struggles to preserve high-frequency geometric details,leading to the loss of salient features.To address this l...The Quadric Error Metrics(QEM)algorithm is a widely used method for mesh simplification;however,it often struggles to preserve high-frequency geometric details,leading to the loss of salient features.To address this limitation,we propose the Salient Feature Sampling Points-based QEM(SFSP-QEM)—also referred to as the Deep Learning-Based Salient Feature-Preserving Algorithm for Mesh Simplification—which incorporates a Salient Feature-Preserving Point Sampler(SFSP).This module leverages deep learning techniques to prioritize the preservation of key geometric features during simplification.Experimental results demonstrate that SFSP-QEM significantly outperforms traditional QEM in preserving geometric details.Specifically,for general models from the Stanford 3D Scanning Repository,which represent typical mesh structures used in mesh simplification benchmarks,the Hausdorff distance of simplified models using SFSP-QEM is reduced by an average of 46.58% compared to those simplified using traditional QEM.In customized models such as the Zigong Lantern used in cultural heritage preservation,SFSP-QEM achieves an average reduction of 28.99% in Hausdorff distance.Moreover,the running time of this method is only 6%longer than that of traditional QEM while significantly improving the preservation of geometric details.These results demonstrate that SFSP-QEMis particularly effective for applications requiring high-fidelity simplification while retaining critical features.展开更多
A novel dual-branch decoding fusion convolutional neural network model(DDFNet)specifically designed for real-time salient object detection(SOD)on steel surfaces is proposed.DDFNet is based on a standard encoder–decod...A novel dual-branch decoding fusion convolutional neural network model(DDFNet)specifically designed for real-time salient object detection(SOD)on steel surfaces is proposed.DDFNet is based on a standard encoder–decoder architecture.DDFNet integrates three key innovations:first,we introduce a novel,lightweight multi-scale progressive aggregation residual network that effectively suppresses background interference and refines defect details,enabling efficient salient feature extraction.Then,we propose an innovative dual-branch decoding fusion structure,comprising the refined defect representation branch and the enhanced defect representation branch,which enhance accuracy in defect region identification and feature representation.Additionally,to further improve the detection of small and complex defects,we incorporate a multi-scale attention fusion module.Experimental results on the public ESDIs-SOD dataset show that DDFNet,with only 3.69 million parameters,achieves detection performance comparable to current state-of-the-art models,demonstrating its potential for real-time industrial applications.Furthermore,our DDFNet-L variant consistently outperforms leading methods in detection performance.The code is available at https://github.com/13140W/DDFNet.展开更多
Video salient object detection(VSOD)aims at locating the most attractive objects in a video by exploring the spatial and temporal features.VSOD poses a challenging task in computer vision,as it involves processing com...Video salient object detection(VSOD)aims at locating the most attractive objects in a video by exploring the spatial and temporal features.VSOD poses a challenging task in computer vision,as it involves processing complex spatial data that is also influenced by temporal dynamics.Despite the progress made in existing VSOD models,they still struggle in scenes of great background diversity within and between frames.Additionally,they encounter difficulties related to accumulated noise and high time consumption during the extraction of temporal features over a long-term duration.We propose a multi-stream temporal enhanced network(MSTENet)to address these problems.It investigates saliency cues collaboration in the spatial domain with a multi-stream structure to deal with the great background diversity challenge.A straightforward,yet efficient approach for temporal feature extraction is developed to avoid the accumulative noises and reduce time consumption.The distinction between MSTENet and other VSOD methods stems from its incorporation of both foreground supervision and background supervision,facilitating enhanced extraction of collaborative saliency cues.Another notable differentiation is the innovative integration of spatial and temporal features,wherein the temporal module is integrated into the multi-stream structure,enabling comprehensive spatial-temporal interactions within an end-to-end framework.Extensive experimental results demonstrate that the proposed method achieves state-of-the-art performance on five benchmark datasets while maintaining a real-time speed of 27 fps(Titan XP).Our code and models are available at https://github.com/RuJiaLe/MSTENet.展开更多
Recently,weak supervision has received growing attention in the field of salient object detection due to the convenience of labelling.However,there is a large performance gap between weakly supervised and fully superv...Recently,weak supervision has received growing attention in the field of salient object detection due to the convenience of labelling.However,there is a large performance gap between weakly supervised and fully supervised salient object detectors because the scribble annotation can only provide very limited foreground/background information.Therefore,an intuitive idea is to infer annotations that cover more complete object and background regions for training.To this end,a label inference strategy is proposed based on the assumption that pixels with similar colours and close positions should have consistent labels.Specifically,k-means clustering algorithm was first performed on both colours and coordinates of original annotations,and then assigned the same labels to points having similar colours with colour cluster centres and near coordinate cluster centres.Next,the same annotations for pixels with similar colours within each kernel neighbourhood was set further.Extensive experiments on six benchmarks demonstrate that our method can significantly improve the performance and achieve the state-of-the-art results.展开更多
A more accurate analysis method on working modes is proposed by considering the winding terminal voltage and the eondueting power device as state parameters. For the three-phase hybrid excitation doubly salient machi...A more accurate analysis method on working modes is proposed by considering the winding terminal voltage and the eondueting power device as state parameters. For the three-phase hybrid excitation doubly salient machine （HEDSM） motor and its three-phase full-bridge inverter, in the proposed analytical method, all possible working modes are generally listed. Then, with the H_PWM-L_ON control strategy, the working modes are detailed with eorresponding equivalent circuits. Experimental results verify the robustness of the analysis.展开更多
The three-phase bridge inverter is used as the converter topology in the power controller for a 9 kW doubly salient permanent magnet （DSPM） motor. Compared with common three-phase bridge inverters, the proposed inve...The three-phase bridge inverter is used as the converter topology in the power controller for a 9 kW doubly salient permanent magnet （DSPM） motor. Compared with common three-phase bridge inverters, the proposed inverter works under more complicated conditions with different principles for special winding back EMFs, position signals of hall sensors, and the given mode of switches. The ideal steady driving principles of the inverter for the motor are given. The working state with asymmetric winding back EMFs, inaccurate position signals of hall sensors, and the changing input voltage is analyzed. Finally, experimental results vertify that the given anal ysis is correct.展开更多
Giora proposed that general principle of salience： salient comprehension of figurative and meanings are processed first and literal language be governed by a more meaning salience determines the type of processing in...Giora proposed that general principle of salience： salient comprehension of figurative and meanings are processed first and literal language be governed by a more meaning salience determines the type of processing invoked. According to the Graded Salience Hypothesis, processing familiar metaphors should involve the activation of both their metaphoric and literal meanings, regardless of the type of context in which they are embedded. Processing less familiar metaphors should activate the literal meaning in both types of contexts; however, in the literally biased context, it should be the only one activated. Processing familiar idioms in context biased towards the idiomatic meaning should evoke their figurative meaning almost exclusively, because their figurative meaning is much more salient than their literal meaning. However, processing less familiar idioms in an idiomatic context should activate both their literal and idiomatic meanings because both meanings enjoy similar salience status.展开更多
A new type of double salient starter/generator is presented, which can be used in aircraft Low Voltage Direct Current (LVDC), Variable Speed Constant Frequency (VSCF) and High Voltage Direct Current (HVDC) systems. Th...A new type of double salient starter/generator is presented, which can be used in aircraft Low Voltage Direct Current (LVDC), Variable Speed Constant Frequency (VSCF) and High Voltage Direct Current (HVDC) systems. The operational theory of the motor and generator is analyzed, and corresponding control strategies are given. An 18kW prototype has been implemented to verify the system performance. It is shown that the DSM S/G system possesses simple structure, high efficiency and flexible control. It is ap...展开更多
The Doubly Salient Electromagnetic Generator(DSEG) is a promising candidate in aircraft generator application due to the simplicity, robustness and reliability. However, the field windings and the armature windings ar...The Doubly Salient Electromagnetic Generator(DSEG) is a promising candidate in aircraft generator application due to the simplicity, robustness and reliability. However, the field windings and the armature windings are strongly coupled, which makes the inductance characteristics non-linear and too complex to model. The complex model with low precision also leads to difficulties in modeling and analysis of the entire aircraft Electrical Power System(EPS). A behavior level modeling method based on modified inductance Support Vector Machine(SVM) is proposed. The Finite Element Analysis(FEA) inductance data are modified based on the experiment results to improve the precision. A functional level modeling method based on input–output characteristics SVM is also proposed. The two modeling methods are applied to a 9 kW DSEG prototype. The steady state and transient process precision of the proposed methods are proved by comparing with the experiment results. Meanwhile, the modeling time consumption, the application time consumption and the calculation resource demand are compared. The DSEG behavior and functional modeling methods provide precious results with high efficiency, which accelerates theoretical analysis and expands the application foreground of the DSEG in the aircraft EPS.展开更多
A new method for automatic salient object segmentation is presented.Salient object segmentation is an important research area in the field of object recognition,image retrieval,image editing,scene reconstruction,and 2...A new method for automatic salient object segmentation is presented.Salient object segmentation is an important research area in the field of object recognition,image retrieval,image editing,scene reconstruction,and 2D/3D conversion.In this work,salient object segmentation is performed using saliency map and color segmentation.Edge,color and intensity feature are extracted from mean shift segmentation(MSS)image,and saliency map is created using these features.First average saliency per segment image is calculated using the color information from MSS image and generated saliency map.Then,second average saliency per segment image is calculated by applying same procedure for the first image to the thresholding,labeling,and hole-filling applied image.Thresholding,labeling and hole-filling are applied to the mean image of the generated two images to get the final salient object segmentation.The effectiveness of proposed method is proved by showing 80%,89%and 80%of precision,recall and F-measure values from the generated salient object segmentation image and ground truth image.展开更多
alient object detection aims at identifying the visually interesting object regions that are consistent with human perception. Multispectral remote sensing images provide rich radiometric information in revealing the ...alient object detection aims at identifying the visually interesting object regions that are consistent with human perception. Multispectral remote sensing images provide rich radiometric information in revealing the physical properties of the observed objects, which leads to great potential to perform salient object detection for remote sensing images. Conventional salient object detection methods often employ handcrafted features to predict saliency by evaluating the pixel-wise or superpixel-wise contrast. With the recent use of deep learning framework, in particular, fully convolutional neural networks, there has been profound progress in visual saliency detection. However, this success has not been extended to multispectral remote sensing images, and existing multispectral salient object detection methods are still mainly based on handcrafted features, essentially due to the difficulties in image acquisition and labeling. In this paper, we propose a novel deep residual network based on a top-down model, which is trained in an end-to-end manner to tackle the above issues in multispectral salient object detection. Our model effectively exploits the saliency cues at different levels of the deep residual network. To overcome the limited availability of remote sensing images in training of our deep residual network, we also introduce a new spectral image reconstruction model that can generate multispectral images from RGB images. Our extensive experimental results using both multispectral and RGB salient object detection datasets demonstrate a significant performance improvement of more than 10% improvement compared with the state-of-the-art methods.展开更多
In traditional analytical method(AM),the magnetic saturation is always ignored to simplify the calculation process.However,synchronous reluctance motors(SynRMs)often operate around saturation point to achieve higher t...In traditional analytical method(AM),the magnetic saturation is always ignored to simplify the calculation process.However,synchronous reluctance motors(SynRMs)often operate around saturation point to achieve higher torque density.Therefore,a new AM is proposed,in which the saturation of stator iron has been considered.The key of the proposed method includes a saturation factor,and an iterative method is adopted to compute the saturation factor in the SynRM by increasing the air-gap length.Especially,the proposed AM can be applied to a SynRM even with shifted-asymmetrical-salient-poles.In the process of AM,the expression of stator magnetomotive force(MMF)is built firstly.Additionally,the air-gap density including slotting effect and salient-poles is calculated.Then,the rotor MMF under saturation of the stator iron is obtained.Therefore,the precision of the instantaneous torque can be improved significantly.Eventually,by the verification of finite elements method(FEM)and experiments,the torque performance of SynRMs with shifted asymmetrical rotor can be predicted accurately by the proposed AM.展开更多
The goal of salient object detection is to estimate the regions which are most likely to attract human's visual attention. As an important image preprocessing procedure to reduce the computational complexity, sali...The goal of salient object detection is to estimate the regions which are most likely to attract human's visual attention. As an important image preprocessing procedure to reduce the computational complexity, salient object detection is still a challenging problem in computer vision. In this paper, we proposed a salient object detection model by integrating local and global superpixel contrast at multiple scales. Three features are computed to estimate the saliency of superpixel. Two optimization measures are utilized to refine the resulting saliency map. Extensive experiments with the state-of-the-art saliency models on four public datasets demonstrate the effectiveness of the proposed model.展开更多
The integrity and fineness characterization of non-connected regions and contours is a major challenge for existing salient object detection.The key to address is how to make full use of the subjective and objective s...The integrity and fineness characterization of non-connected regions and contours is a major challenge for existing salient object detection.The key to address is how to make full use of the subjective and objective structural information obtained in different steps.Therefore,by simulating the human visual mechanism,this paper proposes a novel multi-decoder matching correction network and subjective structural loss.Specifically,the loss pays different attentions to the foreground,boundary,and background of ground truth map in a top-down structure.And the perceived saliency is mapped to the corresponding objective structure of the prediction map,which is extracted in a bottom-up manner.Thus,multi-level salient features can be effectively detected with the loss as constraint.And then,through the mapping of improved binary cross entropy loss,the differences between salient regions and objects are checked to pay attention to the error prone region to achieve excellent error sensitivity.Finally,through tracking the identifying feature horizontally and vertically,the subjective and objective interaction is maximized.Extensive experiments on five benchmark datasets demonstrate that compared with 12 state-of-the-art methods,the algorithm has higher recall and precision,less error and strong robustness and generalization ability,and can predict complete and refined saliency maps.展开更多
Both the XWD1 and XWD2 wells drilled in 2017 in the Wensu salient, northwest Tarim Basin have achieved high-yield industrial oil flow. Based on the comprehensive research on drilling, oil testing, geochemistry and log...Both the XWD1 and XWD2 wells drilled in 2017 in the Wensu salient, northwest Tarim Basin have achieved high-yield industrial oil flow. Based on the comprehensive research on drilling, oil testing, geochemistry and logging data, in combination with the field surveys, 2 D seismic data processing and interpretation as well as sedimentation and accumulation history comparison, we carefully compared the source conditions, migration channels, reservoir-cap distribution and trapping types in the Wensu salient, and subsequently constructed a reservoir-forming pattern. Though the Wensu salient is lack of source rocks, some drainage systems were widely developed and efficiently connected to adjacent fertile depressions. Due to the moderate Miocene paleogeomorphic conditions in the Wensu salient, the delta and shore-shallow lacustrine beach bar sandy bodies were developed within the Jidike formation, and consequently form widely distributed structural-lithologic traps. The hydrocarbon generation, migration and accumulation mainly happened in the Neogene-Quaternary period, which suggests that the reservoir-forming pattern should be characterized as late-period and compound accumulation. It suggests that, although the border belts in the Tarim Basin might be short of source rocks and structural traps, they are potential to accumulate hydrocarbon in a large scale; the description of efficient hydrocarbon migration channels and structural-lithologic traps is crucial for any successful exploration.展开更多
With the improvement of vehicles electrical equipment, the existing silicon rectification generator and permanent magnet generator cannot meet the requirement of the electric power consumption of the modern vehicles e...With the improvement of vehicles electrical equipment, the existing silicon rectification generator and permanent magnet generator cannot meet the requirement of the electric power consumption of the modern vehicles electrical equipment. It is di cult to adjust the air gap magnetic field of the permanent magnet generator. Consequently, the output voltage is not stable. The silicon rectifying generator has the problems of low e ciency and high failure rate.In order to solve these problems, a new type of hybrid excitation generator is developed in this paper. The developed hybrid excitation generator has a double-radial permanent magnet, a salient-pole electromagnetic combined rotor,and a fractional slot winding stator, where each rotor pole corresponds to 4.5 stator teeth. The equivalent magnetic circuit diagram of permanent magnet rotor and magnetic rotor is established. Magnetic field finite element analysis(FEA) software is used to conduct the modeling and simulation analysis on double-radial permanent magnet magnetic field, salient-pole electro-magnetic magnetic field and hybrid magnetic field. The magnetic flux density mold value diagram and vector diagram are obtained. The diagrams are used to verify the feasibility of this design. The designed electromagnetic coupling regulator controller can ensure the stable voltage export by changing the magnitude and direction of the excitation current to adjust the size of the air gap magnetic field. Therefore, the problem of output voltage instability in the wide speed range and wide load range of the hybrid excitation generator is solved.展开更多
The correct rate of detection for fabric defect is affected by low contrast of images. Aiming at the problem,frequencytuned salient map is used to detect the fabric defect. Firstly,the images of fabric defect are divi...The correct rate of detection for fabric defect is affected by low contrast of images. Aiming at the problem,frequencytuned salient map is used to detect the fabric defect. Firstly,the images of fabric defect are divided into blocks. Then,the blocks are highlighted by frequency-tuned salient algorithm. Simultaneously,gray-level co-occurrence matrix is used to extract the characteristic value of each rectangular patch. Finally,PNN is used to detect the defect on the fabric image. The performance of proposed algorithm is estimated off-line by two sets of fabric defect images. The theoretical argument is supported by experimental results.展开更多
The National Program on the Development of Chinese Women （2011-2020）, herein- after referred to as the ＂NewProgram,＂ was published in August 2011. This is an important document designed to ensure implementation of...The National Program on the Development of Chinese Women （2011-2020）, herein- after referred to as the ＂NewProgram,＂ was published in August 2011. This is an important document designed to ensure implementation of the basic state strategy of gender equal- ity and the all-round development of Chinese women. The New Program is a part of China＇s policy program for the protection of human rights. It sets 57 major targets to be attained over the decade. The targets cover seven fields, namely, health, education, economy,展开更多
Gait recognition aims to identify individuals by distinguishing unique walking patterns based on video-level pedestrian silhouettes.Previous studies have focused on designing powerful feature extractors to model the s...Gait recognition aims to identify individuals by distinguishing unique walking patterns based on video-level pedestrian silhouettes.Previous studies have focused on designing powerful feature extractors to model the spatio-temporal dependencies of gait,thereby obtaining gait features that contain rich semantic information.However,they have overlooked the potential of feature maps to construct discriminative gait embeddings.In this work,we propose a novel model,EmbedGait,which is designed to learn salient gait embeddings for improved recognition results.Specifically,our framework starts with a frame-level spatial alignment to maintain inter-sequence consistency.Then,horizontal salient mapping(HSM)module is designed to extract the representative embeddings and discard the background information by a designed pooling operation.The subsequent adaptive embedding weighting(AEW)module is used to adaptively highlight the salient embeddings of different body parts and channels.Extensive experiments on the Gait3D,GREW and SUSTech1K datasets demonstrate that our approach improves comparable performance in several benchmarks tests.For example,our proposed EmbedGait achieves rank-1 accuracies of 77.3%,79.0%and 79.6%on Gait3D,GREW and SUSTech1K,respectively.展开更多
文摘At present, salient object detection (SOD) has achieved considerable progress. However, the methods that perform well still face the issue of inadequate detection accuracy. For example, sometimes there are problems of missed and false detections. Effectively optimizing features to capture key information and better integrating different levels of features to enhance their complementarity are two significant challenges in the domain of SOD. In response to these challenges, this study proposes a novel SOD method based on multi-strategy feature optimization. We propose the multi-size feature extraction module (MSFEM), which uses the attention mechanism, the multi-level feature fusion, and the residual block to obtain finer features. This module provides robust support for the subsequent accurate detection of the salient object. In addition, we use two rounds of feature fusion and the feedback mechanism to optimize the features obtained by the MSFEM to improve detection accuracy. The first round of feature fusion is applied to integrate the features extracted by the MSFEM to obtain more refined features. Subsequently, the feedback mechanism and the second round of feature fusion are applied to refine the features, thereby providing a stronger foundation for accurately detecting salient objects. To improve the fusion effect, we propose the feature enhancement module (FEM) and the feature optimization module (FOM). The FEM integrates the upper and lower features with the optimized features obtained by the FOM to enhance feature complementarity. The FOM uses different receptive fields, the attention mechanism, and the residual block to more effectively capture key information. Experimental results demonstrate that our method outperforms 10 state-of-the-art SOD methods.
基金Our research was funded by the Sichuan Key Provincial Research Base of Intelligent Tourism(No.ZHZJ23-02)supported by the Scientific Research and Innovation Team Program of Sichuan University of Science and Engineering(No.SUSE652A006)+1 种基金Additional support was provided by the National Cultural and Tourism Science and Technology Innovation Research andDevelopment Project(No.202417)the Lantern Culture and Crafts Innovation Key Laboratory Project of the Sichuan ProvincialDepartment of Culture and Tourism(No.SCWLCD-A02).
文摘The Quadric Error Metrics(QEM)algorithm is a widely used method for mesh simplification;however,it often struggles to preserve high-frequency geometric details,leading to the loss of salient features.To address this limitation,we propose the Salient Feature Sampling Points-based QEM(SFSP-QEM)—also referred to as the Deep Learning-Based Salient Feature-Preserving Algorithm for Mesh Simplification—which incorporates a Salient Feature-Preserving Point Sampler(SFSP).This module leverages deep learning techniques to prioritize the preservation of key geometric features during simplification.Experimental results demonstrate that SFSP-QEM significantly outperforms traditional QEM in preserving geometric details.Specifically,for general models from the Stanford 3D Scanning Repository,which represent typical mesh structures used in mesh simplification benchmarks,the Hausdorff distance of simplified models using SFSP-QEM is reduced by an average of 46.58% compared to those simplified using traditional QEM.In customized models such as the Zigong Lantern used in cultural heritage preservation,SFSP-QEM achieves an average reduction of 28.99% in Hausdorff distance.Moreover,the running time of this method is only 6%longer than that of traditional QEM while significantly improving the preservation of geometric details.These results demonstrate that SFSP-QEMis particularly effective for applications requiring high-fidelity simplification while retaining critical features.
基金supported in part by the National Key R&D Program of China(Grant No.2023YFB3307604)the Shanxi Province Basic Research Program Youth Science Research Project(Grant Nos.202303021212054 and 202303021212046)+3 种基金the Key Projects Supported by Hebei Natural Science Foundation(Grant No.E2024203125)the National Science Foundation of China(Grant No.52105391)the Hebei Provincial Science and Technology Major Project(Grant No.23280101Z)the National Key Laboratory of Metal Forming Technology and Heavy Equipment Open Fund(Grant No.S2308100.W17).
文摘A novel dual-branch decoding fusion convolutional neural network model(DDFNet)specifically designed for real-time salient object detection(SOD)on steel surfaces is proposed.DDFNet is based on a standard encoder–decoder architecture.DDFNet integrates three key innovations:first,we introduce a novel,lightweight multi-scale progressive aggregation residual network that effectively suppresses background interference and refines defect details,enabling efficient salient feature extraction.Then,we propose an innovative dual-branch decoding fusion structure,comprising the refined defect representation branch and the enhanced defect representation branch,which enhance accuracy in defect region identification and feature representation.Additionally,to further improve the detection of small and complex defects,we incorporate a multi-scale attention fusion module.Experimental results on the public ESDIs-SOD dataset show that DDFNet,with only 3.69 million parameters,achieves detection performance comparable to current state-of-the-art models,demonstrating its potential for real-time industrial applications.Furthermore,our DDFNet-L variant consistently outperforms leading methods in detection performance.The code is available at https://github.com/13140W/DDFNet.
基金funded by the Natural Science Foundation China(NSFC)under Grant No.62203192.
文摘Video salient object detection(VSOD)aims at locating the most attractive objects in a video by exploring the spatial and temporal features.VSOD poses a challenging task in computer vision,as it involves processing complex spatial data that is also influenced by temporal dynamics.Despite the progress made in existing VSOD models,they still struggle in scenes of great background diversity within and between frames.Additionally,they encounter difficulties related to accumulated noise and high time consumption during the extraction of temporal features over a long-term duration.We propose a multi-stream temporal enhanced network(MSTENet)to address these problems.It investigates saliency cues collaboration in the spatial domain with a multi-stream structure to deal with the great background diversity challenge.A straightforward,yet efficient approach for temporal feature extraction is developed to avoid the accumulative noises and reduce time consumption.The distinction between MSTENet and other VSOD methods stems from its incorporation of both foreground supervision and background supervision,facilitating enhanced extraction of collaborative saliency cues.Another notable differentiation is the innovative integration of spatial and temporal features,wherein the temporal module is integrated into the multi-stream structure,enabling comprehensive spatial-temporal interactions within an end-to-end framework.Extensive experimental results demonstrate that the proposed method achieves state-of-the-art performance on five benchmark datasets while maintaining a real-time speed of 27 fps(Titan XP).Our code and models are available at https://github.com/RuJiaLe/MSTENet.
文摘Recently,weak supervision has received growing attention in the field of salient object detection due to the convenience of labelling.However,there is a large performance gap between weakly supervised and fully supervised salient object detectors because the scribble annotation can only provide very limited foreground/background information.Therefore,an intuitive idea is to infer annotations that cover more complete object and background regions for training.To this end,a label inference strategy is proposed based on the assumption that pixels with similar colours and close positions should have consistent labels.Specifically,k-means clustering algorithm was first performed on both colours and coordinates of original annotations,and then assigned the same labels to points having similar colours with colour cluster centres and near coordinate cluster centres.Next,the same annotations for pixels with similar colours within each kernel neighbourhood was set further.Extensive experiments on six benchmarks demonstrate that our method can significantly improve the performance and achieve the state-of-the-art results.
文摘A more accurate analysis method on working modes is proposed by considering the winding terminal voltage and the eondueting power device as state parameters. For the three-phase hybrid excitation doubly salient machine （HEDSM） motor and its three-phase full-bridge inverter, in the proposed analytical method, all possible working modes are generally listed. Then, with the H_PWM-L_ON control strategy, the working modes are detailed with eorresponding equivalent circuits. Experimental results verify the robustness of the analysis.
文摘The three-phase bridge inverter is used as the converter topology in the power controller for a 9 kW doubly salient permanent magnet （DSPM） motor. Compared with common three-phase bridge inverters, the proposed inverter works under more complicated conditions with different principles for special winding back EMFs, position signals of hall sensors, and the given mode of switches. The ideal steady driving principles of the inverter for the motor are given. The working state with asymmetric winding back EMFs, inaccurate position signals of hall sensors, and the changing input voltage is analyzed. Finally, experimental results vertify that the given anal ysis is correct.
文摘Giora proposed that general principle of salience： salient comprehension of figurative and meanings are processed first and literal language be governed by a more meaning salience determines the type of processing invoked. According to the Graded Salience Hypothesis, processing familiar metaphors should involve the activation of both their metaphoric and literal meanings, regardless of the type of context in which they are embedded. Processing less familiar metaphors should activate the literal meaning in both types of contexts; however, in the literally biased context, it should be the only one activated. Processing familiar idioms in context biased towards the idiomatic meaning should evoke their figurative meaning almost exclusively, because their figurative meaning is much more salient than their literal meaning. However, processing less familiar idioms in an idiomatic context should activate both their literal and idiomatic meanings because both meanings enjoy similar salience status.
文摘A new type of double salient starter/generator is presented, which can be used in aircraft Low Voltage Direct Current (LVDC), Variable Speed Constant Frequency (VSCF) and High Voltage Direct Current (HVDC) systems. The operational theory of the motor and generator is analyzed, and corresponding control strategies are given. An 18kW prototype has been implemented to verify the system performance. It is shown that the DSM S/G system possesses simple structure, high efficiency and flexible control. It is ap...
基金co-supported by the National Natural Science Foundation for Outstanding Young Scholar of China (No. 51622704)Jiangsu Provincial Science Funds for Distinguished Young Scientists of China (No. BK20150033)Postgraduate Research & Practice Innovation Program of Jiangsu Province (No. KYLX16_0358)
文摘The Doubly Salient Electromagnetic Generator(DSEG) is a promising candidate in aircraft generator application due to the simplicity, robustness and reliability. However, the field windings and the armature windings are strongly coupled, which makes the inductance characteristics non-linear and too complex to model. The complex model with low precision also leads to difficulties in modeling and analysis of the entire aircraft Electrical Power System(EPS). A behavior level modeling method based on modified inductance Support Vector Machine(SVM) is proposed. The Finite Element Analysis(FEA) inductance data are modified based on the experiment results to improve the precision. A functional level modeling method based on input–output characteristics SVM is also proposed. The two modeling methods are applied to a 9 kW DSEG prototype. The steady state and transient process precision of the proposed methods are proved by comparing with the experiment results. Meanwhile, the modeling time consumption, the application time consumption and the calculation resource demand are compared. The DSEG behavior and functional modeling methods provide precious results with high efficiency, which accelerates theoretical analysis and expands the application foreground of the DSEG in the aircraft EPS.
文摘A new method for automatic salient object segmentation is presented.Salient object segmentation is an important research area in the field of object recognition,image retrieval,image editing,scene reconstruction,and 2D/3D conversion.In this work,salient object segmentation is performed using saliency map and color segmentation.Edge,color and intensity feature are extracted from mean shift segmentation(MSS)image,and saliency map is created using these features.First average saliency per segment image is calculated using the color information from MSS image and generated saliency map.Then,second average saliency per segment image is calculated by applying same procedure for the first image to the thresholding,labeling,and hole-filling applied image.Thresholding,labeling and hole-filling are applied to the mean image of the generated two images to get the final salient object segmentation.The effectiveness of proposed method is proved by showing 80%,89%and 80%of precision,recall and F-measure values from the generated salient object segmentation image and ground truth image.
基金National 1000 Young Talents Plan of ChinaNational Natural Science Foundation of China(61420106007,61671387,61871325)DECRA of Australica Resenrch Council (DE140100180).
文摘alient object detection aims at identifying the visually interesting object regions that are consistent with human perception. Multispectral remote sensing images provide rich radiometric information in revealing the physical properties of the observed objects, which leads to great potential to perform salient object detection for remote sensing images. Conventional salient object detection methods often employ handcrafted features to predict saliency by evaluating the pixel-wise or superpixel-wise contrast. With the recent use of deep learning framework, in particular, fully convolutional neural networks, there has been profound progress in visual saliency detection. However, this success has not been extended to multispectral remote sensing images, and existing multispectral salient object detection methods are still mainly based on handcrafted features, essentially due to the difficulties in image acquisition and labeling. In this paper, we propose a novel deep residual network based on a top-down model, which is trained in an end-to-end manner to tackle the above issues in multispectral salient object detection. Our model effectively exploits the saliency cues at different levels of the deep residual network. To overcome the limited availability of remote sensing images in training of our deep residual network, we also introduce a new spectral image reconstruction model that can generate multispectral images from RGB images. Our extensive experimental results using both multispectral and RGB salient object detection datasets demonstrate a significant performance improvement of more than 10% improvement compared with the state-of-the-art methods.
基金This work was supported in part by the National Natural Science Foundation of China(51707083)in part by the Natural Science Foundation of Jiangsu Province(BK20190848)+1 种基金in part by the China Postdoctoral Science Foundation(2019M661746)by the Priority Academic Program Development of Jiangsu Higher Education Institutions.
文摘In traditional analytical method(AM),the magnetic saturation is always ignored to simplify the calculation process.However,synchronous reluctance motors(SynRMs)often operate around saturation point to achieve higher torque density.Therefore,a new AM is proposed,in which the saturation of stator iron has been considered.The key of the proposed method includes a saturation factor,and an iterative method is adopted to compute the saturation factor in the SynRM by increasing the air-gap length.Especially,the proposed AM can be applied to a SynRM even with shifted-asymmetrical-salient-poles.In the process of AM,the expression of stator magnetomotive force(MMF)is built firstly.Additionally,the air-gap density including slotting effect and salient-poles is calculated.Then,the rotor MMF under saturation of the stator iron is obtained.Therefore,the precision of the instantaneous torque can be improved significantly.Eventually,by the verification of finite elements method(FEM)and experiments,the torque performance of SynRMs with shifted asymmetrical rotor can be predicted accurately by the proposed AM.
基金the Natural Science Foundation of China(Nos.61602349,61375053,and 61273225)the China Scholarship Council(No.201508420248)Hubei Chengguang Talented Youth Development Foundation(No.2015B22)
文摘The goal of salient object detection is to estimate the regions which are most likely to attract human's visual attention. As an important image preprocessing procedure to reduce the computational complexity, salient object detection is still a challenging problem in computer vision. In this paper, we proposed a salient object detection model by integrating local and global superpixel contrast at multiple scales. Three features are computed to estimate the saliency of superpixel. Two optimization measures are utilized to refine the resulting saliency map. Extensive experiments with the state-of-the-art saliency models on four public datasets demonstrate the effectiveness of the proposed model.
基金supported by the National Natural Science Foundation of China(No.52174021)Key Research and Develop-ment Project of Hainan Province(No.ZDYF2022GXJS 003).
文摘The integrity and fineness characterization of non-connected regions and contours is a major challenge for existing salient object detection.The key to address is how to make full use of the subjective and objective structural information obtained in different steps.Therefore,by simulating the human visual mechanism,this paper proposes a novel multi-decoder matching correction network and subjective structural loss.Specifically,the loss pays different attentions to the foreground,boundary,and background of ground truth map in a top-down structure.And the perceived saliency is mapped to the corresponding objective structure of the prediction map,which is extracted in a bottom-up manner.Thus,multi-level salient features can be effectively detected with the loss as constraint.And then,through the mapping of improved binary cross entropy loss,the differences between salient regions and objects are checked to pay attention to the error prone region to achieve excellent error sensitivity.Finally,through tracking the identifying feature horizontally and vertically,the subjective and objective interaction is maximized.Extensive experiments on five benchmark datasets demonstrate that compared with 12 state-of-the-art methods,the algorithm has higher recall and precision,less error and strong robustness and generalization ability,and can predict complete and refined saliency maps.
基金Supported by China Geological Survey Projects(DD20160203,DD20189602,DD20160169)
文摘Both the XWD1 and XWD2 wells drilled in 2017 in the Wensu salient, northwest Tarim Basin have achieved high-yield industrial oil flow. Based on the comprehensive research on drilling, oil testing, geochemistry and logging data, in combination with the field surveys, 2 D seismic data processing and interpretation as well as sedimentation and accumulation history comparison, we carefully compared the source conditions, migration channels, reservoir-cap distribution and trapping types in the Wensu salient, and subsequently constructed a reservoir-forming pattern. Though the Wensu salient is lack of source rocks, some drainage systems were widely developed and efficiently connected to adjacent fertile depressions. Due to the moderate Miocene paleogeomorphic conditions in the Wensu salient, the delta and shore-shallow lacustrine beach bar sandy bodies were developed within the Jidike formation, and consequently form widely distributed structural-lithologic traps. The hydrocarbon generation, migration and accumulation mainly happened in the Neogene-Quaternary period, which suggests that the reservoir-forming pattern should be characterized as late-period and compound accumulation. It suggests that, although the border belts in the Tarim Basin might be short of source rocks and structural traps, they are potential to accumulate hydrocarbon in a large scale; the description of efficient hydrocarbon migration channels and structural-lithologic traps is crucial for any successful exploration.
基金Supported by National Natural Science Foundation of China(Grant No.51507096)Shandong Provincial Natural Science Foundation of China(Grant No.ZR2014JL035)
文摘With the improvement of vehicles electrical equipment, the existing silicon rectification generator and permanent magnet generator cannot meet the requirement of the electric power consumption of the modern vehicles electrical equipment. It is di cult to adjust the air gap magnetic field of the permanent magnet generator. Consequently, the output voltage is not stable. The silicon rectifying generator has the problems of low e ciency and high failure rate.In order to solve these problems, a new type of hybrid excitation generator is developed in this paper. The developed hybrid excitation generator has a double-radial permanent magnet, a salient-pole electromagnetic combined rotor,and a fractional slot winding stator, where each rotor pole corresponds to 4.5 stator teeth. The equivalent magnetic circuit diagram of permanent magnet rotor and magnetic rotor is established. Magnetic field finite element analysis(FEA) software is used to conduct the modeling and simulation analysis on double-radial permanent magnet magnetic field, salient-pole electro-magnetic magnetic field and hybrid magnetic field. The magnetic flux density mold value diagram and vector diagram are obtained. The diagrams are used to verify the feasibility of this design. The designed electromagnetic coupling regulator controller can ensure the stable voltage export by changing the magnitude and direction of the excitation current to adjust the size of the air gap magnetic field. Therefore, the problem of output voltage instability in the wide speed range and wide load range of the hybrid excitation generator is solved.
文摘The correct rate of detection for fabric defect is affected by low contrast of images. Aiming at the problem,frequencytuned salient map is used to detect the fabric defect. Firstly,the images of fabric defect are divided into blocks. Then,the blocks are highlighted by frequency-tuned salient algorithm. Simultaneously,gray-level co-occurrence matrix is used to extract the characteristic value of each rectangular patch. Finally,PNN is used to detect the defect on the fabric image. The performance of proposed algorithm is estimated off-line by two sets of fabric defect images. The theoretical argument is supported by experimental results.
文摘The National Program on the Development of Chinese Women （2011-2020）, herein- after referred to as the ＂NewProgram,＂ was published in August 2011. This is an important document designed to ensure implementation of the basic state strategy of gender equal- ity and the all-round development of Chinese women. The New Program is a part of China＇s policy program for the protection of human rights. It sets 57 major targets to be attained over the decade. The targets cover seven fields, namely, health, education, economy,
基金supported by the International Science and Technology Cooperation Project of Guangzhou Economic and Technological Development District(No.2023GH16).
文摘Gait recognition aims to identify individuals by distinguishing unique walking patterns based on video-level pedestrian silhouettes.Previous studies have focused on designing powerful feature extractors to model the spatio-temporal dependencies of gait,thereby obtaining gait features that contain rich semantic information.However,they have overlooked the potential of feature maps to construct discriminative gait embeddings.In this work,we propose a novel model,EmbedGait,which is designed to learn salient gait embeddings for improved recognition results.Specifically,our framework starts with a frame-level spatial alignment to maintain inter-sequence consistency.Then,horizontal salient mapping(HSM)module is designed to extract the representative embeddings and discard the background information by a designed pooling operation.The subsequent adaptive embedding weighting(AEW)module is used to adaptively highlight the salient embeddings of different body parts and channels.Extensive experiments on the Gait3D,GREW and SUSTech1K datasets demonstrate that our approach improves comparable performance in several benchmarks tests.For example,our proposed EmbedGait achieves rank-1 accuracies of 77.3%,79.0%and 79.6%on Gait3D,GREW and SUSTech1K,respectively.
