Exploring visual dictionaries: A model driven perspective

Good representative dictionaries is the most critical part of the BoVW: Bag of Visual Words scheme, used for such tasks as category identification. The paradigm of learning dictionaries from datasets is by far the most widely used approach and there exists a plethora of methods to this effect. Dictionary learning methods demand abundant data, and when the amount of training data is limited, the quality of dictionaries and consequently the performance of BoVW methods suffer. A much less explored path for creating visual dictionaries starts from the knowledge of primitives in appearance models and creates families of parametric shape models. In this work, we develop shape models starting from a small number of primitives and develop a visual dictionary using various nonlinear operations and nonlinear combinations. Compared with the existing model-driven schemes, our method is able to represent and characterize images in various image understanding applications with competitive, and often better performance.     

Multi-view representation learning for multi-view action recognition

Although multiple methods have been proposed for human action recognition, the existing multi-view approaches cannot well discover meaningful relationship among multiple action categories from different views. To handle this problem, this paper proposes an multi-view learning approach for multi-view action recognition. First, the proposed method leverages the popular visual representation method, bag-of-visual-words (BoVW)/fisher vector (FV), to represent individual videos in each view. Second, the sparse coding algorithm is utilized to transfer the low-level features of various views into the discriminative and high-level semantics space. Third, we employ the multi-task learning (MTL) approach for joint action modeling and discovery of latent relationship among different action categories. The extensive experimental results on M²I and IXMAS datasets have demonstrated the effectiveness of our proposed approach. Moreover, the experiments further demonstrate that the discovered latent relationship can benefit multi-view model learning to augment the performance of action recognition.     

Improving retrieval framework using information gain models

Content-based image retrieval systems are meant to retrieve the most similar images of a collection to a query image. One of the most well-known models widely applied for this task is the bag of visual words (BoVW) model. In this paper, we introduce a study of different information gain models used for the construction of a visual vocabulary. In the proposed framework, information gain models are used as a discriminative information to index image features and select the ones that have the highest information gain values. We introduce some extensions to further improve the performance of the proposed framework: mixing different vocabularies and extending the BoVW to bag of visual phrases. Exhaustive experiments show the interest of information gain models on our retrieval framework.     

联合多尺度多特征的高分遥感图像场景分类

高分辨率遥感图像地物信息丰富,但场景构成复杂,目前基于手工设计的特征提取方法不能满足复杂场景分类的需求,而非监督特征学习方法尽管能够挖掘局部图像块的本征结构,但单一种类及尺度的特征难以有效表达实际应用中复杂遥感场景特性,导致分类性能受限.针对此问题,本文提出了一种基于多尺度多特征的遥感场景分类方法.该算法首先设计了一种改进的谱聚类非监督特征（iUFL-SC）以有效表征图像块的本征结构,然后通过密集采样提取每幅遥感场景的iUFL-SC、LBP、SIFT等三种多尺度局部图像块特征,并通过视觉词袋模型（BoVW）获得场景的中层特征表达,以实现更为准确详实的特征描述,最后基于直方图交叉核的支持向量机（HIKSVM）进行分类.在UC Merced数据集以及WHU-RS19数据集上的实验结果表明本文方法可对遥感场景进行鉴别特征提取,有效提高分类性能.     

一种基于随机化视觉词典组和查询扩展的目标检索方法

在目标检索领域,当前主流的解决方案是视觉词典法(Bag of Visual Words, BoVW),然而,传统的BoVW方法具有时间效率低、内存消耗大以及视觉单词同义性和歧义性的问题。针对以上问题,该文提出了一种基于随机化视觉词典组和查询扩展的目标检索方法。首先,该方法采用精确欧氏位置敏感哈希(Exact Euclidean Locality Sensitive Hashing, E2LSH)对训练图像库的局部特征点进行聚类,生成一组支持动态扩充的随机化视觉词典组;然后,基于这组词典构建视觉词汇分布直方图和索引文件;最后,引入一种查询扩展策略完成目标检索。实验结果表明,与传统方法相比,该文方法有效地增强了目标对象的可区分性,能够较大地提高目标检索精度,同时,对大规模数据库有较好的适用性。     

Spatially Enhanced Bags of Visual Words Representation to Improve Traffic Signs Recognition

Traffic signs play a very vital role in safe driving and in avoiding accidents by informing the driver about the speed limits or possible dangers such as icy roads, imminent road works or pedestrian crossings. Considering the processing time and classification accuracy as a whole, a novel approach for visual words construction was presented, which takes the spatial information of keypoints into account in order to enhance the quality of visual words generated from extracted keypoints using the distance and angle information in the Bags of Visual Words (BoVW) representation. In this paper, we proposed a new computationally efficient method to model global spatial distribution of visual words by taking into consideration the spatial relationships of its visual words. In the first step, the region of interest is extracted using a scanning window with a Haar cascade detector and an AdaBoost classifier to reduce the computational region in the hypothesis generation step. Second, the regions are represented with BoVW and spatial information for classification. Experimental results show that the suggested method could reach comparable performance of the state-of-the-art approaches with less computational complexity and shorter training time. It clearly demonstrates the complementarity of the additional relative spatial information provided by our approach to improve accuracy while maintaining short retrieval time, and can obtain a better traffic sign recognition accuracy than the methods based on the traditional BoVW model.     

一种新颖的多agent强化学习方法

提出了一种综合了模块化结构、利益分配学习以及对手建模技术的多agent强化学习方法,利用模块化学习结构来克服状态空间的维数灾问题,将Q-学习与利益分配学习相结合以加快学习速度,采用基于观察的对手建模来预测其他agent的动作分布.追捕问题的仿真结果验证了所提方法的有效性.     

基于域与样例平衡的多源迁移学习方法

针对如何有效使用多源域的决策知识去预测目标域样例标签的问题，提出一种平衡域与样例信息的多源迁移学习算法.为实现上述目的，本文提出了一种基于域与样例平衡的多源迁移学习方法（Multi-source Transfer Learning by Balancing both Domains and Instances，MTL-BDI）.该方法的基本思想是将域层面和样例层面的双加权平衡项嵌入到迁移学习的原始目标函数中，然后利用交替优化技术对提出的目标函数进行有效求解.在文本和图像数据集上的大量实验表明，该方法在分类精度方面确实优于现有的多源迁移学习方法MCC-SVM（Multiple Convex Combination of SVM）、A-SVM（Adaptive SVM）、Multi-KMM（Multiple Kernel Mean Matching）和DAM（Domain Adaptation Machine）.     

一种基于自适应软分配的图像分类方法

视觉词袋模型(BoVW)是当前图像分类领域的主流方法,然而,视觉单词同义性和歧义性问题严重制约了该模型的性能,进而降低图像分类准确率。针对该问题,本文提出一种基于自适应软分配的图像分类方法。该方法首先对尺度不变特征变换(SIFT)特征映射到视觉单词的距离进行分析,按一定的规则进行归类,并针对具有不同模糊程度的SIFT特征采用自适应的分配策略;然后,通过卡方模型分析各个视觉单词与图像类别之间的相关性,并依此去除视觉停用词(VSW),重构视觉单词统计直方图;最后,输入到支持向量机(SVM)完成分类。实验结果表明,该优化方法能有效地降低视觉单词同义性和歧义性问题带来的影响,增强视觉单词的区分性,进而提高图像分类准确率。     

Relevance feedback using generalized Bayesian framework with region-based optimization learning.

This paper presents a generalized Bayesian framework for relevance feedback in content-based image retrieval. The proposed feedback technique is based on the Bayesian learning method and incorporates a time-varying user model into the formulation. We define the user model with two terms: a target query and a user conception. The target query is aimed to learn the common features from relevant images so as to specify the user's ideal query. The user conception is aimed to learn a parameter set to determine the time-varying matching criterion. Therefore, at each feedback step, the learning process updates not only the target distribution, but also the target query and the matching criterion. In addition, another objective of this paper is to conduct the relevance feedback on images represented in region level. We formulate the matching criterion using a weighting scheme and proposed a region clustering technique to determine the region correspondence between relevant images. With the proposed region clustering technique, we derive a representation in region level to characterize the target query. Experiments demonstrate that the proposed method combined with time-varying user model indeed achieves satisfactory results and our proposed region-based techniques further improve the retrieval accuracy.     

基于多尺度上下文语义信息的图像场景分类算法

传统视觉词典模型没有考虑图像的多尺度和上下文语义共生关系.本文提出一种基于多尺度上下文语义信息的图像场景分类算法.首先,对图像进行多尺度分解,从多个尺度提取不同粒度的视觉信息;其次利用基于密度的自适应选择算法确定最优概率潜在语义分析模型主题数;然后,结合Markov随机场共同挖掘图像块的上下文语义共生信息,得到图像的多尺度直方图表示;最后结合支持向量机实现场景分类.实验结果表明,本文算法能有效利用图像的多尺度和上下文语义信息,提高视觉单词的语义准确性,从而改善场景分类性能.     

PSO-based learning of sub-band adaptive thresholding function for image denoising

In this paper, a new computationally efficient approach has been proposed for denoising the images which are corrupted by Gaussian noise. In this approach, relatively recent category of stochastic global optimization technique i.e., particle swarm optimization (PSO) technique have been proposed for learning the parameters of adaptive thresholding function required for optimum performance. The proposed PSO-based denoising approach not only speeds up the optimization but also improves the performance in comparison with wavelet transform-based thresholding neural network (WT-TNN) approach. The results obtained shows better edge preservation performance with bior6.8 wavelet filter when compared to db8 wavelet filter. Further, problem of dependency of learning time on initial value of thresholding parameters and noise level in the image have been sorted out in the proposed approach.     

基于低秩表示的鲁棒判别特征子空间学习模型

特征子空间学习是图像识别及分类任务的关键技术之一,传统的特征子空间学习模型面临两个主要的问题。一方面是如何使样本在投影到特征空间后有效地保持其局部结构和判别性。另一方面是当样本含噪时传统学习模型所发生的失效问题。针对上述两个问题,该文提出一种基于低秩表示(LRR)的判别特征子空间学习模型,该模型的主要贡献包括:通过低秩表示探究样本的局部结构,并利用表示系数作为样本在投影空间的相似性约束,使投影子空间能够更好地保持样本的局部近邻关系;为提高模型的抗噪能力,构造了一种利用低秩重构样本的判别特征学习约束项,同时增强模型的判别性和鲁棒性;设计了一种基于交替优化技术的迭代数值求解方案来保证算法的收敛性。该文在多个视觉数据集上进行分类任务的对比实验,实验结果表明所提算法在分类准确度和鲁棒性方面均优于传统特征学习方法。     

Proposal of a voice activity detector for noise reduction

In some noise reduction techniques, the learning of noise characteristics is required and a voice activity detector (VAD) must be used to determine noise sequences. The VAD proposed is derived from the coherence function computed from two microphones. It is combined with a noise reduction technique and its influence evaluated     

Anomaly detection of smart metering system for power management with battery storage system/electric vehicle

A novel smart metering technique capable of anomaly detection was proposed for real-time home power management system. Smart meter data generated in real-time were obtained from 900 households of single apartments. To detect outliers and missing values in smart meter data, a deep learning model, the autoencoder, consisting of a graph convolutional network and bidirectional long short-term memory network, was applied to the smart metering technique. Power management based on the smart metering technique was executed by multi-objective optimization in the presence of a battery storage system and an electric vehicle. The results of the power management employing the proposed smart metering technique indicate a reduction in electricity cost and amount of power supplied by the grid compared to the results of power management without anomaly detection.     

Compensatory fuzzy neural networks-based intelligent detection ofabnormal neonatal cerebral Doppler ultrasound waveforms

Compensatory fuzzy neural networks (CFNN) without normalization, which can be trained with a backpropagation learning algorithm, are proposed as a pattern recognition technique for the intelligent detection of Doppler ultrasound waveforms of abnormal neonatal cerebral hemodynamics. Doppler ultrasound signals were recorded from the anterior cerebral arteries of 40 normal full-term babies and 14 mature babies with intracranial pathology. The features of normal and abnormal groups as inputs to the pattern recognition algorithms were extracted from the maximum-velocity waveforms by using principal component analysis. The proposed technique is compared with the CFNN with normalization and other pattern recognition techniques applied to Doppler ultrasound signals from various arteries. The results show that the proposed method is superior to the other techniques, and can be a powerful way to analyze Doppler ultrasound signals from various arteries     

Disguised‐Face Discriminator for Embedded Systems

In this paper, we introduce an improved adaptive boosting (AdaBoost) classifier and its application, a disguised‐face discriminator that discriminates between bare and disguised faces. The proposed classifier is based on an AdaBoost learning algorithm and regression technique. In the process, the lookup table of AdaBoost learning is utilized. The proposed method is verified on the captured images under several real environments. Experimental results and analysis show the proposed method has a higher and faster performance than other well‐known methods.     

Nonlinear System Control Using Adaptive Neural Fuzzy Networks Based on a Modified Differential Evolution

This study presents an adaptive neural fuzzy network (ANFN) controller based on a modified differential evolution (MODE) for solving control problems. The proposed ANFN controller adopts a functional link neural network as the consequent part of the fuzzy rules. Thus, the consequent part of the ANFN controller is a nonlinear combination of input variables. The proposed MODE learning algorithm adopts an evolutionary learning method to optimize the controller parameters. For design optimization, a new criterion is introduced. A hardware-in-the loop control technique is developed and applied to the designed ANFN controller using the MODE learning algorithm. The proposed ANFN controller with the MODE learning algorithm (ANFN-MODE) is used in two practical applications—the planetary-train-type inverted pendulum system and the magnetic levitation system. The experiment is developed in a real-time visual simulation environment. Experimental results of this study have demonstrated the robustness and effectiveness of the proposed ANFN-MODE controller.      

A novel approach to multiagent reinforcement learning: utilizing OLAP mining in the learning process

Reinforcement learning is considered as a strong method for learning in multiagent systems environments. However, it still has some drawbacks, including modeling other learning agents present in the domain as part of the state of the environment, and some states are much less experienced than others, or some state-action pairs are never visited during the learning phase. Further, before the learning process is completed, an agent cannot exhibit a certain behavior in some states that may be sufficiently experienced. This shows that learning in a partially observable and dynamic multiagent systems environment still constitutes a difficult and major research problem that is worth further investigation. Motivated by this, in this paper, a novel learning approach that integrates online analytical processing (OLAP)-based data mining into the process is proposed. First, a data cube OLAP architecture that facilitates effective storage and processing of the state information reported by agents is described. This way, the action of the other agent, even one not in the visual environment of the agent under consideration, can simply be estimated by extracting online association rules, a well-known data mining technique, from the constructed data cube. Then, a new action selection model that is also based on association rules mining is presented. Finally, states that are not sufficiently experienced by mining multiple-level association rules from the proposed data cube are generalized. Experiments conducted on a well-known pursuit domain show the robustness and effectiveness of the proposed learning approach.