In this paper,we propose hierarchical attention dual network(DNet)for fine-grained image classification.The DNet can randomly select pairs of inputs from the dataset and compare the differences between them through hi...In this paper,we propose hierarchical attention dual network(DNet)for fine-grained image classification.The DNet can randomly select pairs of inputs from the dataset and compare the differences between them through hierarchical attention feature learning,which are used simultaneously to remove noise and retain salient features.In the loss function,it considers the losses of difference in paired images according to the intra-variance and inter-variance.In addition,we also collect the disaster scene dataset from remote sensing images and apply the proposed method to disaster scene classification,which contains complex scenes and multiple types of disasters.Compared to other methods,experimental results show that the DNet with hierarchical attention is robust to different datasets and performs better.展开更多
Deep learning has achieved excellent results in various tasks in the field of computer vision,especially in fine-grained visual categorization.It aims to distinguish the subordinate categories of the label-level categ...Deep learning has achieved excellent results in various tasks in the field of computer vision,especially in fine-grained visual categorization.It aims to distinguish the subordinate categories of the label-level categories.Due to high intra-class variances and high inter-class similarity,the fine-grained visual categorization is extremely challenging.This paper first briefly introduces and analyzes the related public datasets.After that,some of the latest methods are reviewed.Based on the feature types,the feature processing methods,and the overall structure used in the model,we divide them into three types of methods:methods based on general convolutional neural network(CNN)and strong supervision of parts,methods based on single feature processing,and meth-ods based on multiple feature processing.Most methods of the first type have a relatively simple structure,which is the result of the initial research.The methods of the other two types include models that have special structures and training processes,which are helpful to obtain discriminative features.We conduct a specific analysis on several methods with high accuracy on pub-lic datasets.In addition,we support that the focus of the future research is to solve the demand of existing methods for the large amount of the data and the computing power.In terms of tech-nology,the extraction of the subtle feature information with the burgeoning vision transformer(ViT)network is also an important research direction.展开更多
The initiating behavior of fine-grained explosives by small flyer is studied. The diameter of small flyer in this device is 1mm. The test results indicate that the granularity of explosives has great effect on its fly...The initiating behavior of fine-grained explosives by small flyer is studied. The diameter of small flyer in this device is 1mm. The test results indicate that the granularity of explosives has great effect on its flyer initiating sensitivity.The flyer initiating sensitivity of the fine-grained explosives is higher and the critical initiating energy is lower than that of common explosives. For common explosive, the flyer initiating sensitivity increases as the density is reduced. But for the fine-grained explosive, the test results are exactly opposite.展开更多
人脸特征蕴含诸多信息,在面部属性和情感分析任务中具有重要价值,而面部特征的多样性和复杂性使人脸分析任务变得困难。针对上述难题,从面部细粒度特征角度出发,提出基于上下文通道注意力机制的人脸属性估计和表情识别(FAER)模型。首先...人脸特征蕴含诸多信息,在面部属性和情感分析任务中具有重要价值,而面部特征的多样性和复杂性使人脸分析任务变得困难。针对上述难题,从面部细粒度特征角度出发,提出基于上下文通道注意力机制的人脸属性估计和表情识别(FAER)模型。首先,构建基于ConvNext的局部特征编码骨干网络,并运用骨干网络编码局部特征的有效性来充分表征人脸局部特征之间的差异性;其次,提出上下文通道注意力(CC Attention)机制,通过动态自适应调整特征通道上的权重信息,表征深度特征的全局和局部特征,从而弥补骨干网络编码全局特征能力的不足;最后,设计不同分类策略,针对人脸属性估计(FAE)和面部表情识别(FER)任务,分别采用不同损失函数组合,以促使模型学习更多的面部细粒度特征。实验结果表明,所提FAER模型在人脸属性数据集CelebA(CelebFaces Attributes)上取得了91.87%的平均准确率,相较于次优模型SwinFace(Swin transformer for Face)高出0.55个百分点;在面部表情数据集RAF-DB和AffectNet上分别取得了91.75%和66.66%的准确率,相较于次优模型TransFER(Transformers for Facial Expression Recognition)分别高出0.84和0.43个百分点。展开更多
服装已经成为网络购物的重要商品之一,实现精准的符合用户个性化审美的服装推荐系统,已经成为热门研究内容。针对提取用户的细粒度兴趣特征不全面,导致推荐系统的准确性低问题,提出融合长短期偏好的服装推荐算法;针对数据稀疏以及数据...服装已经成为网络购物的重要商品之一,实现精准的符合用户个性化审美的服装推荐系统,已经成为热门研究内容。针对提取用户的细粒度兴趣特征不全面,导致推荐系统的准确性低问题,提出融合长短期偏好的服装推荐算法;针对数据稀疏以及数据单一性,导致推荐结果个性化、多样性低的问题,利用跨模态数据和注意力机制使模型学习出更为精准的差异性用户特征。在真实数据集Clothing Shoes and Jewelry上,将所设计的模型(PCR)与经典的循环神经网络RNN、基于矩阵分解MF-BPR模型以及改进的矩阵分解TARMF模型进行性能比对,PCR模型在关键性能评价指标NDCG、Precision@K和Recall@K均有提升。实验结果表明该模型在服装推荐系统中是可行与有效的。展开更多
基金Supported by the National Natural Science Foundation of China(61601176)。
文摘In this paper,we propose hierarchical attention dual network(DNet)for fine-grained image classification.The DNet can randomly select pairs of inputs from the dataset and compare the differences between them through hierarchical attention feature learning,which are used simultaneously to remove noise and retain salient features.In the loss function,it considers the losses of difference in paired images according to the intra-variance and inter-variance.In addition,we also collect the disaster scene dataset from remote sensing images and apply the proposed method to disaster scene classification,which contains complex scenes and multiple types of disasters.Compared to other methods,experimental results show that the DNet with hierarchical attention is robust to different datasets and performs better.
基金supported by the National Natural Science Foundation of China(61571453,61806218).
文摘Deep learning has achieved excellent results in various tasks in the field of computer vision,especially in fine-grained visual categorization.It aims to distinguish the subordinate categories of the label-level categories.Due to high intra-class variances and high inter-class similarity,the fine-grained visual categorization is extremely challenging.This paper first briefly introduces and analyzes the related public datasets.After that,some of the latest methods are reviewed.Based on the feature types,the feature processing methods,and the overall structure used in the model,we divide them into three types of methods:methods based on general convolutional neural network(CNN)and strong supervision of parts,methods based on single feature processing,and meth-ods based on multiple feature processing.Most methods of the first type have a relatively simple structure,which is the result of the initial research.The methods of the other two types include models that have special structures and training processes,which are helpful to obtain discriminative features.We conduct a specific analysis on several methods with high accuracy on pub-lic datasets.In addition,we support that the focus of the future research is to solve the demand of existing methods for the large amount of the data and the computing power.In terms of tech-nology,the extraction of the subtle feature information with the burgeoning vision transformer(ViT)network is also an important research direction.
基金the Nature Science Foundation of Shanxi Province (20021064)
文摘The initiating behavior of fine-grained explosives by small flyer is studied. The diameter of small flyer in this device is 1mm. The test results indicate that the granularity of explosives has great effect on its flyer initiating sensitivity.The flyer initiating sensitivity of the fine-grained explosives is higher and the critical initiating energy is lower than that of common explosives. For common explosive, the flyer initiating sensitivity increases as the density is reduced. But for the fine-grained explosive, the test results are exactly opposite.
文摘人脸特征蕴含诸多信息,在面部属性和情感分析任务中具有重要价值,而面部特征的多样性和复杂性使人脸分析任务变得困难。针对上述难题,从面部细粒度特征角度出发,提出基于上下文通道注意力机制的人脸属性估计和表情识别(FAER)模型。首先,构建基于ConvNext的局部特征编码骨干网络,并运用骨干网络编码局部特征的有效性来充分表征人脸局部特征之间的差异性;其次,提出上下文通道注意力(CC Attention)机制,通过动态自适应调整特征通道上的权重信息,表征深度特征的全局和局部特征,从而弥补骨干网络编码全局特征能力的不足;最后,设计不同分类策略,针对人脸属性估计(FAE)和面部表情识别(FER)任务,分别采用不同损失函数组合,以促使模型学习更多的面部细粒度特征。实验结果表明,所提FAER模型在人脸属性数据集CelebA(CelebFaces Attributes)上取得了91.87%的平均准确率,相较于次优模型SwinFace(Swin transformer for Face)高出0.55个百分点;在面部表情数据集RAF-DB和AffectNet上分别取得了91.75%和66.66%的准确率,相较于次优模型TransFER(Transformers for Facial Expression Recognition)分别高出0.84和0.43个百分点。
文摘服装已经成为网络购物的重要商品之一,实现精准的符合用户个性化审美的服装推荐系统,已经成为热门研究内容。针对提取用户的细粒度兴趣特征不全面,导致推荐系统的准确性低问题,提出融合长短期偏好的服装推荐算法;针对数据稀疏以及数据单一性,导致推荐结果个性化、多样性低的问题,利用跨模态数据和注意力机制使模型学习出更为精准的差异性用户特征。在真实数据集Clothing Shoes and Jewelry上,将所设计的模型(PCR)与经典的循环神经网络RNN、基于矩阵分解MF-BPR模型以及改进的矩阵分解TARMF模型进行性能比对,PCR模型在关键性能评价指标NDCG、Precision@K和Recall@K均有提升。实验结果表明该模型在服装推荐系统中是可行与有效的。
