Recent state-of-the-art semi-supervised learning(SSL)methods usually use data augmentations as core components.Such methods,however,are limited to simple transformations such as the augmentations under the instance’s...Recent state-of-the-art semi-supervised learning(SSL)methods usually use data augmentations as core components.Such methods,however,are limited to simple transformations such as the augmentations under the instance’s naive representations or the augmentations under the instance’s semantic representations.To tackle this problem,we offer a unique insight into data augmentations and propose a novel data-augmentation-based semi-supervised learning method,called Attentive Neighborhood Feature Aug-mentation(ANFA).The motivation of our method lies in the observation that the relationship between the given feature and its neighborhood may contribute to constructing more reliable transformations for the data,and further facilitating the classifier to distinguish the ambiguous features from the low-dense regions.Specially,we first project the labeled and unlabeled data points into an embedding space and then construct a neighbor graph that serves as a similarity measure based on the similar representations in the embedding space.Then,we employ an attention mechanism to transform the target features into augmented ones based on the neighbor graph.Finally,we formulate a novel semi-supervised loss by encouraging the predictions of the interpolations of augmented features to be consistent with the corresponding interpolations of the predictions of the target features.We carried out exper-iments on SVHN and CIFAR-10 benchmark datasets and the experimental results demonstrate that our method outperforms the state-of-the-art methods when the number of labeled examples is limited.展开更多
Background:While nest attentiveness plays a critical role in the reproductive success of avian species,nest attentiveness data with high temporal resolution is not available for many species.However,improvements in bo...Background:While nest attentiveness plays a critical role in the reproductive success of avian species,nest attentiveness data with high temporal resolution is not available for many species.However,improvements in both video monitoring and temperature logging devices present an opportunity to increase our understanding of this aspect of avian behavior.Methods:To investigate nest attentiveness behaviors and evaluate these technologies,we monitored 13 nests across two Common Tern(Sterna hirundo)breeding colonies with a paired video camera-temperature logger approach,while monitoring 63 additional nests with temperature loggers alone.Observations occurred from May to August of 2017 on Poplar(Chesapeake Bay,Maryland,USA)and Skimmer Islands(Isle of Wight Bay,Maryland,USA).We examined data respective to four times of day:Morning(civil dawn‒11:59),Peak(12:00‒16:00),Cooling(16:01‒civil dusk),and Night(civil dusk‒civil dawn).Results:While successful nests had mostly short duration off-bouts and maintained consistent nest attentiveness throughout the day,failed nests had dramatic reductions in nest attentiveness during the Cooling and Night periods(p<0.05)with one colony experiencing repeated nocturnal abandonment due to predation pressure from a Great Horned Owl(Bubo virginianus).Incubation appeared to ameliorate ambient temperatures during Night,as nests were significantly warmer during Night when birds were on versus off the nest(p<0.05).Meanwhile,off-bouts during the Peak period occurred during higher ambient temperatures,perhaps due to adults leaving the nest during the hottest periods to perform belly soaking.Unfortunately,temperature logger data alone had limited ability to predict nest attentiveness status during shorter bouts,with results highly dependent on time of day and bout duration.While our methods did not affect hatching success(p>0.05),video-monitored nests did have significantly lower clutch sizes(p<0.05).Conclusions:The paired use of iButtons and video cameras enabled a detailed description of the incubation behavior of COTE.However,while promising for future research,the logistical and potential biological complications involved in the use of these methods suggest that careful planning is needed before these devices are utilized to ensure data is collected in a safe and successful manner.展开更多
Health indicator(HI)construction is a crucial task in degradation evaluation and facilitates the prognostic and health management(PHM)of rotating machinery.Excluding interference from artificial labeling,the HI constr...Health indicator(HI)construction is a crucial task in degradation evaluation and facilitates the prognostic and health management(PHM)of rotating machinery.Excluding interference from artificial labeling,the HI construction approaches in an unsupervised manner have attracted substantial attention.Nevertheless,current unsupervised methods generally struggle with two problems:(1)ignorance of both redundancy between features and global variability of features during the feature selection process;(2)inadequate utilization of information from different sampling moments.To tackle these problems,this work develops a novel unsupervised approach for HI construction that integrates multi-criterion feature selection and the Attentive Variational Autoencoder(Attentive VAE).Explicitly,a multi-criterion feature selection(Mc FS)algorithm together with an elaborately designed metric is proposed to determine a superior feature subset,considering the relevance,the redundancy,and the global variability of features simultaneously.Then,for the adequate utilization of the information from distinct sampling moments,a deep learning model named Attentive VAE is established.The Attentive VAE is solely fed with the selected features in the health state for model training and the HI is derived through the reconstruction error to reveal the degradation degree of machinery.Two case studies based on genuine experimental datasets are involved to quantitatively evaluate the superiority of the developed approach,demonstrating its superiority over other unsupervised methods for characterizing degradation processes.The effectiveness of both the Mc FS algorithm and the Attentive VAE is verified by ablation experiments,respectively.展开更多
The utilization of visual attention enhances the performance of image classification tasks.Previous attentionbased models have demonstrated notable performance,but many of these models exhibit reduced accuracy when co...The utilization of visual attention enhances the performance of image classification tasks.Previous attentionbased models have demonstrated notable performance,but many of these models exhibit reduced accuracy when confronted with inter-class and intra-class similarities and differences.Neural-Controlled Differential Equations(N-CDE’s)and Neural Ordinary Differential Equations(NODE’s)are extensively utilized within this context.NCDE’s possesses the capacity to effectively illustrate both inter-class and intra-class similarities and differences with enhanced clarity.To this end,an attentive neural network has been proposed to generate attention maps,which uses two different types of N-CDE’s,one for adopting hidden layers and the other to generate attention values.Two distinct attention techniques are implemented including time-wise attention,also referred to as bottom N-CDE’s;and element-wise attention,called topN-CDE’s.Additionally,a trainingmethodology is proposed to guarantee that the training problem is sufficiently presented.Two classification tasks including fine-grained visual classification andmulti-label classification,are utilized to evaluate the proposedmodel.The proposedmethodology is employed on five publicly available datasets,including CUB-200-2011,ImageNet-1K,PASCAL VOC 2007,PASCAL VOC 2012,and MS COCO.The obtained visualizations have demonstrated that N-CDE’s are better appropriate for attention-based activities in comparison to conventional NODE’s.展开更多
In recent years,learning-based low-light image enhancement methods have shown excellent performance,but the heuristic design adopted by most methods requires high engineering skills for developers,causing expensive in...In recent years,learning-based low-light image enhancement methods have shown excellent performance,but the heuristic design adopted by most methods requires high engineering skills for developers,causing expensive inference costs that are unfriendly to the hardware platform.To handle this issue,we propose to automatically discover an efficient architecture,called progressive attentive Retinex network(PAR-Net).We define a new attentive Retinex framework by introducing the attention mechanism to strengthen structural representation.A multi-level search space containing micro-level on the operation and macro-level on the cell is established to realize meticulous construction.To endow the searched architecture with the hardware-aware property,we develop a latency-constrained progressive search strategy that successfully improves the model capability by explicitly expressing the intrinsic relationship between different models defined in the attentive Retinex framework.Extensive quantitative and qualitative experimental results fully justify the superiority of our proposed approach against other state-of-the-art methods.A series of analytical evaluations is performed to illustrate the validity of our proposed algorithm.展开更多
针对分布式电源和新型负荷容量累积造成负荷影响因素多元化和不确定性特性增强的问题,文中提出一种采用记忆神经网络和曲线形状修正的负荷预测方法。在负荷峰值预测中,采用最大信息系数计算负荷峰值与影响因素的非线性相关性,实现对输...针对分布式电源和新型负荷容量累积造成负荷影响因素多元化和不确定性特性增强的问题,文中提出一种采用记忆神经网络和曲线形状修正的负荷预测方法。在负荷峰值预测中,采用最大信息系数计算负荷峰值与影响因素的非线性相关性,实现对输入特征的筛选;综合考虑负荷峰值序列的长短期自相关性和输入特征与负荷峰值的不同程度相关性,结合Attention机制和双向长短时记忆(bidirectional long short-term memory,BiLSTM)神经网络建立负荷峰值预测模型。在负荷标幺曲线预测中,通过误差倒数法组合相似日和相邻日,建立负荷标幺曲线预测模型;针对预测偏差的非平稳特征,利用自适应噪声的完全集成经验模态分解和BiLSTM网络建立误差预测模型,对曲线形状进行修正。应用中国北方某城市的区域电网负荷数据为算例,验证了所提模型的有效性。展开更多
针对现有的数字化档案多标签分类方法存在分类标签之间缺少关联性的问题,提出一种用于档案多标签分类的深层神经网络模型ALBERT-Seq2Seq-Attention.该模型通过ALBERT(A Little BERT)预训练语言模型内部多层双向的Transfomer结构获取进...针对现有的数字化档案多标签分类方法存在分类标签之间缺少关联性的问题,提出一种用于档案多标签分类的深层神经网络模型ALBERT-Seq2Seq-Attention.该模型通过ALBERT(A Little BERT)预训练语言模型内部多层双向的Transfomer结构获取进行文本特征向量的提取,并获得上下文语义信息;将预训练提取的文本特征作为Seq2Seq-Attention(Sequence to Sequence-Attention)模型的输入序列,构建标签字典以获取多标签间的关联关系.将分类模型在3种数据集上分别进行对比实验,结果表明:模型分类的效果F1值均超过90%.该模型不仅能提高档案文本的多标签分类效果,也能关注标签之间的相关关系.展开更多
Recently,deep learning-based image inpainting methods have made great strides in reconstructing damaged regions.However,these methods often struggle to produce satisfactory results when dealing with missing images wit...Recently,deep learning-based image inpainting methods have made great strides in reconstructing damaged regions.However,these methods often struggle to produce satisfactory results when dealing with missing images with large holes,leading to distortions in the structure and blurring of textures.To address these problems,we combine the advantages of transformers and convolutions to propose an image inpainting method that incorporates edge priors and attention mechanisms.The proposed method aims to improve the results of inpainting large holes in images by enhancing the accuracy of structure restoration and the ability to recover texture details.This method divides the inpainting task into two phases:edge prediction and image inpainting.Specifically,in the edge prediction phase,a transformer architecture is designed to combine axial attention with standard self-attention.This design enhances the extraction capability of global structural features and location awareness.It also balances the complexity of self-attention operations,resulting in accurate prediction of the edge structure in the defective region.In the image inpainting phase,a multi-scale fusion attention module is introduced.This module makes full use of multi-level distant features and enhances local pixel continuity,thereby significantly improving the quality of image inpainting.To evaluate the performance of our method.comparative experiments are conducted on several datasets,including CelebA,Places2,and Facade.Quantitative experiments show that our method outperforms the other mainstream methods.Specifically,it improves Peak Signal-to-Noise Ratio(PSNR)and Structure Similarity Index Measure(SSIM)by 1.141~3.234 db and 0.083~0.235,respectively.Moreover,it reduces Learning Perceptual Image Patch Similarity(LPIPS)and Mean Absolute Error(MAE)by 0.0347~0.1753 and 0.0104~0.0402,respectively.Qualitative experiments reveal that our method excels at reconstructing images with complete structural information and clear texture details.Furthermore,our model exhibits impressive performance in terms of the number of parameters,memory cost,and testing time.展开更多
基金supported by the National Natural Science Foundation of China (Nos.62072127,62002076,61906049)Natural Science Foundation of Guangdong Province (Nos.2023A1515011774,2020A1515010423)+4 种基金Project 6142111180404 supported by CNKLSTISS,Science and Technology Program of Guangzhou,China (No.202002030131)Guangdong basic and applied basic research fund joint fund Youth Fund (No.2019A1515110213)Open Fund Project of Fujian Provincial Key Laboratory of Information Processing and Intelligent Control (Minjiang University) (No.MJUKF-IPIC202101)Natural Science Foundation of Guangdong Province No.2020A1515010423)Scientific research project for Guangzhou University (No.RP2022003).
文摘Recent state-of-the-art semi-supervised learning(SSL)methods usually use data augmentations as core components.Such methods,however,are limited to simple transformations such as the augmentations under the instance’s naive representations or the augmentations under the instance’s semantic representations.To tackle this problem,we offer a unique insight into data augmentations and propose a novel data-augmentation-based semi-supervised learning method,called Attentive Neighborhood Feature Aug-mentation(ANFA).The motivation of our method lies in the observation that the relationship between the given feature and its neighborhood may contribute to constructing more reliable transformations for the data,and further facilitating the classifier to distinguish the ambiguous features from the low-dense regions.Specially,we first project the labeled and unlabeled data points into an embedding space and then construct a neighbor graph that serves as a similarity measure based on the similar representations in the embedding space.Then,we employ an attention mechanism to transform the target features into augmented ones based on the neighbor graph.Finally,we formulate a novel semi-supervised loss by encouraging the predictions of the interpolations of augmented features to be consistent with the corresponding interpolations of the predictions of the target features.We carried out exper-iments on SVHN and CIFAR-10 benchmark datasets and the experimental results demonstrate that our method outperforms the state-of-the-art methods when the number of labeled examples is limited.
基金This work was supported by the U.S.Army Corps of Engineers(Baltimore District),U.S.Geological Survey(Patuxent Wildlife Research Center)the University of Maryland,the Maryland Department of Natural Resources(Wildlife and Heritage Program)the Maryland Environmental Service,and the Maryland Coastal Bays Program.
文摘Background:While nest attentiveness plays a critical role in the reproductive success of avian species,nest attentiveness data with high temporal resolution is not available for many species.However,improvements in both video monitoring and temperature logging devices present an opportunity to increase our understanding of this aspect of avian behavior.Methods:To investigate nest attentiveness behaviors and evaluate these technologies,we monitored 13 nests across two Common Tern(Sterna hirundo)breeding colonies with a paired video camera-temperature logger approach,while monitoring 63 additional nests with temperature loggers alone.Observations occurred from May to August of 2017 on Poplar(Chesapeake Bay,Maryland,USA)and Skimmer Islands(Isle of Wight Bay,Maryland,USA).We examined data respective to four times of day:Morning(civil dawn‒11:59),Peak(12:00‒16:00),Cooling(16:01‒civil dusk),and Night(civil dusk‒civil dawn).Results:While successful nests had mostly short duration off-bouts and maintained consistent nest attentiveness throughout the day,failed nests had dramatic reductions in nest attentiveness during the Cooling and Night periods(p<0.05)with one colony experiencing repeated nocturnal abandonment due to predation pressure from a Great Horned Owl(Bubo virginianus).Incubation appeared to ameliorate ambient temperatures during Night,as nests were significantly warmer during Night when birds were on versus off the nest(p<0.05).Meanwhile,off-bouts during the Peak period occurred during higher ambient temperatures,perhaps due to adults leaving the nest during the hottest periods to perform belly soaking.Unfortunately,temperature logger data alone had limited ability to predict nest attentiveness status during shorter bouts,with results highly dependent on time of day and bout duration.While our methods did not affect hatching success(p>0.05),video-monitored nests did have significantly lower clutch sizes(p<0.05).Conclusions:The paired use of iButtons and video cameras enabled a detailed description of the incubation behavior of COTE.However,while promising for future research,the logistical and potential biological complications involved in the use of these methods suggest that careful planning is needed before these devices are utilized to ensure data is collected in a safe and successful manner.
基金supported by the National Key Research and Development Program of China(Grant No.2021YFB3400700)the China Academy of Railway Sciences Corporation Limited within the major issues of the fund(Grant No.2021YJ212)+1 种基金the National Natural Science Foundation of China(Grant Nos.12072188,12121002)the Natural Science Foundation of Shanghai(Grant No.20ZR1425200)。
文摘Health indicator(HI)construction is a crucial task in degradation evaluation and facilitates the prognostic and health management(PHM)of rotating machinery.Excluding interference from artificial labeling,the HI construction approaches in an unsupervised manner have attracted substantial attention.Nevertheless,current unsupervised methods generally struggle with two problems:(1)ignorance of both redundancy between features and global variability of features during the feature selection process;(2)inadequate utilization of information from different sampling moments.To tackle these problems,this work develops a novel unsupervised approach for HI construction that integrates multi-criterion feature selection and the Attentive Variational Autoencoder(Attentive VAE).Explicitly,a multi-criterion feature selection(Mc FS)algorithm together with an elaborately designed metric is proposed to determine a superior feature subset,considering the relevance,the redundancy,and the global variability of features simultaneously.Then,for the adequate utilization of the information from distinct sampling moments,a deep learning model named Attentive VAE is established.The Attentive VAE is solely fed with the selected features in the health state for model training and the HI is derived through the reconstruction error to reveal the degradation degree of machinery.Two case studies based on genuine experimental datasets are involved to quantitatively evaluate the superiority of the developed approach,demonstrating its superiority over other unsupervised methods for characterizing degradation processes.The effectiveness of both the Mc FS algorithm and the Attentive VAE is verified by ablation experiments,respectively.
基金Institutional Fund Projects under Grant No.(IFPIP:638-830-1443).
文摘The utilization of visual attention enhances the performance of image classification tasks.Previous attentionbased models have demonstrated notable performance,but many of these models exhibit reduced accuracy when confronted with inter-class and intra-class similarities and differences.Neural-Controlled Differential Equations(N-CDE’s)and Neural Ordinary Differential Equations(NODE’s)are extensively utilized within this context.NCDE’s possesses the capacity to effectively illustrate both inter-class and intra-class similarities and differences with enhanced clarity.To this end,an attentive neural network has been proposed to generate attention maps,which uses two different types of N-CDE’s,one for adopting hidden layers and the other to generate attention values.Two distinct attention techniques are implemented including time-wise attention,also referred to as bottom N-CDE’s;and element-wise attention,called topN-CDE’s.Additionally,a trainingmethodology is proposed to guarantee that the training problem is sufficiently presented.Two classification tasks including fine-grained visual classification andmulti-label classification,are utilized to evaluate the proposedmodel.The proposedmethodology is employed on five publicly available datasets,including CUB-200-2011,ImageNet-1K,PASCAL VOC 2007,PASCAL VOC 2012,and MS COCO.The obtained visualizations have demonstrated that N-CDE’s are better appropriate for attention-based activities in comparison to conventional NODE’s.
文摘In recent years,learning-based low-light image enhancement methods have shown excellent performance,but the heuristic design adopted by most methods requires high engineering skills for developers,causing expensive inference costs that are unfriendly to the hardware platform.To handle this issue,we propose to automatically discover an efficient architecture,called progressive attentive Retinex network(PAR-Net).We define a new attentive Retinex framework by introducing the attention mechanism to strengthen structural representation.A multi-level search space containing micro-level on the operation and macro-level on the cell is established to realize meticulous construction.To endow the searched architecture with the hardware-aware property,we develop a latency-constrained progressive search strategy that successfully improves the model capability by explicitly expressing the intrinsic relationship between different models defined in the attentive Retinex framework.Extensive quantitative and qualitative experimental results fully justify the superiority of our proposed approach against other state-of-the-art methods.A series of analytical evaluations is performed to illustrate the validity of our proposed algorithm.
文摘针对分布式电源和新型负荷容量累积造成负荷影响因素多元化和不确定性特性增强的问题,文中提出一种采用记忆神经网络和曲线形状修正的负荷预测方法。在负荷峰值预测中,采用最大信息系数计算负荷峰值与影响因素的非线性相关性,实现对输入特征的筛选;综合考虑负荷峰值序列的长短期自相关性和输入特征与负荷峰值的不同程度相关性,结合Attention机制和双向长短时记忆(bidirectional long short-term memory,BiLSTM)神经网络建立负荷峰值预测模型。在负荷标幺曲线预测中,通过误差倒数法组合相似日和相邻日,建立负荷标幺曲线预测模型;针对预测偏差的非平稳特征,利用自适应噪声的完全集成经验模态分解和BiLSTM网络建立误差预测模型,对曲线形状进行修正。应用中国北方某城市的区域电网负荷数据为算例,验证了所提模型的有效性。
文摘针对现有的数字化档案多标签分类方法存在分类标签之间缺少关联性的问题,提出一种用于档案多标签分类的深层神经网络模型ALBERT-Seq2Seq-Attention.该模型通过ALBERT(A Little BERT)预训练语言模型内部多层双向的Transfomer结构获取进行文本特征向量的提取,并获得上下文语义信息;将预训练提取的文本特征作为Seq2Seq-Attention(Sequence to Sequence-Attention)模型的输入序列,构建标签字典以获取多标签间的关联关系.将分类模型在3种数据集上分别进行对比实验,结果表明:模型分类的效果F1值均超过90%.该模型不仅能提高档案文本的多标签分类效果,也能关注标签之间的相关关系.
基金supported in part by the National Natural Science Foundation of China under Grant 62062061/in part by the Major Project Cultivation Fund of Xizang Minzu University under Grant 324112300447.
文摘Recently,deep learning-based image inpainting methods have made great strides in reconstructing damaged regions.However,these methods often struggle to produce satisfactory results when dealing with missing images with large holes,leading to distortions in the structure and blurring of textures.To address these problems,we combine the advantages of transformers and convolutions to propose an image inpainting method that incorporates edge priors and attention mechanisms.The proposed method aims to improve the results of inpainting large holes in images by enhancing the accuracy of structure restoration and the ability to recover texture details.This method divides the inpainting task into two phases:edge prediction and image inpainting.Specifically,in the edge prediction phase,a transformer architecture is designed to combine axial attention with standard self-attention.This design enhances the extraction capability of global structural features and location awareness.It also balances the complexity of self-attention operations,resulting in accurate prediction of the edge structure in the defective region.In the image inpainting phase,a multi-scale fusion attention module is introduced.This module makes full use of multi-level distant features and enhances local pixel continuity,thereby significantly improving the quality of image inpainting.To evaluate the performance of our method.comparative experiments are conducted on several datasets,including CelebA,Places2,and Facade.Quantitative experiments show that our method outperforms the other mainstream methods.Specifically,it improves Peak Signal-to-Noise Ratio(PSNR)and Structure Similarity Index Measure(SSIM)by 1.141~3.234 db and 0.083~0.235,respectively.Moreover,it reduces Learning Perceptual Image Patch Similarity(LPIPS)and Mean Absolute Error(MAE)by 0.0347~0.1753 and 0.0104~0.0402,respectively.Qualitative experiments reveal that our method excels at reconstructing images with complete structural information and clear texture details.Furthermore,our model exhibits impressive performance in terms of the number of parameters,memory cost,and testing time.
