基于混合神经网络的脑电时空特征情感分类

提出一种脑电图（electroencephalograph,简称EEG）数据表示方法,将一维链式EEG向量序列转换成二维网状矩阵序列,使矩阵结构与EEG电极位置的脑区分布相对应,以此来更好地表示物理上多个相邻电极EEG信号之间的空间相关性.再应用滑动窗将二维矩阵序列分成一个个等长的时间片段,作为新的融合了EEG时空相关性的数据表示.还提出了级联卷积-循环神经网络（CASC_CNN_LSTM）与级联卷积-卷积神经网络（CASC_CNN_CNN）这两种混合深度学习模型,二者都通过CNN卷积神经网络从转换的二维网状EEG数据表示中捕获物理上相邻脑电信号之间的空间相关性,而前者通过LSTM循环神经网络学习EEG数据流在时序上的依赖关系,后者则通过CNN卷积神经网络挖掘局部时间与空间更深层的相关判别性特征,从而精确识别脑电信号中包含的情感类别.在大规模脑电数据集DEAP上进行被试内效价维度上两类情感分类实验,结果显示,所提出的CASC_CNN_LSTM和CASC_CNN_CNN网络在二维网状EEG时空特征上的平均分类准确率分别达到93.15%和92.37%,均高于基准模型和现有最新方法的性能,表明该模型有效提高了EEG情感识别的准确率和鲁棒性,可以有效地应用到基于EEG的情感分类与识别相关应用中.     

相对剥夺感对大学生抑郁的影响:链式中介效应分析

目的 探讨相对剥夺感对大学生抑郁的影响机制.方法 采用相对剥夺感问卷、情绪调节策略问卷、领悟社会支持量表和流调中心抑郁量表对799名大学生进行调查.结果(1)相对剥夺感、认知重评、表达抑制、领悟社会支持与抑郁两两均呈显著相关;(2)相对剥夺感可以通过五条路径影响大学生抑郁:认知重评、表达抑制和领悟社会支持的独立中介作用...     

融合邻域回归和稀疏表示的图像超分辨率重构

对于图像超分辨率重建而言,通常会将图像的整体信息作为研究对象.然而图像本身含有的大量结构信息并没有得到充分利用.为了提高超分辨率重建的效果,实现对不同特征信息的利用,提出了一种融合邻域回归和稀疏表示的图像超分辨率重构算法.依据图像所具有的低秩性对高分辨率图像进行分解,获得高分辨率图像的低秩部分和稀疏部分;将对应的低分辨...     

基于卷积与稀疏编码的半监督学习方法

卷积神经网络(CNN)在半监督学习中取得了良好的成绩,其在训练阶段既利用有标记样本,也利用无标记样本帮助规范化学习模型。为进一步加强半监督模型的特征学习能力,提高其在图像分类时的性能表现,本文提出一种联合深度半监督卷积神经网络和字典学习的端到端半监督学习方法,称为Semi-supervised Learning based on Sparse Coding and Convolution(SSSConv);该算法框架旨在学习到鉴别性更强的图像特征表示。SSSConv首先利用CNN提取特征,并对所提取特征进行正交投影变换,下一步通过学习其稀疏编码的低维嵌入以得到图像的特征表示,最后据此进行分类。整个模型框架可进行端到端的半监督学习训练,CNN提取特征部分和稀疏编码字典学习部分具有统一的损失函数,目标一致。本文利用共轭梯度下降算法、链式法则和反向传播等算法对目标函数的参数进行优化,将稀疏编码的相关参数约束于流形上,CNN参数既可定义在欧氏空间,也可以进一步定义在正交空间中。基于半监督分类任务的实验结果验证了所提出SSSConv框架的有效性,与现有方法相比具有较强的竞争力。     

A cross‐cultural comparison of colour emotion for two‐colour combinations

Psychophysical experiments were conducted in the UK, Taiwan, France, Germany, Spain, Sweden, Argentina, and Iran to assess colour emotion for two‐colour combinations using semantic scales warm/cool, heavy/light, active/passive, and like/dislike. A total of 223 observers participated, each presented with 190 colour pairs as the stimuli, shown individually on a cathode ray tube display. The results show consistent responses across cultures only for warm/cool, heavy/light, and active/passive. The like/dislike scale, however, showed some differences between the observer groups, in particular between the Argentinian responses and those obtained from the other observers. Factor analysis reveals that the Argentinian observers preferred passive colour pairs to active ones more than the other observers. In addition to the cultural difference in like/dislike, the experimental results show some effects of gender, professional background (design vs. nondesign), and age. Female observers were found to prefer colour pairs with high‐lightness or low‐chroma values more than their male counterparts. Observers with a design background liked low‐chroma colour pairs or those containing colours of similar hue more than nondesign observers. Older observers liked colour pairs with high‐lightness or high‐chroma values more than young observers did. Based on the findings, a two‐level theory of colour emotion is proposed, in which warm/cool, heavy/light, and active/passive are identified as the reactive‐level responses and like/dislike the reflective‐level response. © 2010 Wiley Periodicals, Inc. Col Res Appl, 2012     

Request for existing experimental datasets on colour emotion and colour harmony

The Technical Committee 1‐86 of the International Commission on Illumination on “Models of colour emotion and harmony” is requesting the submission of datasets for use in developing new models of colour emotion and colour harmony. The data should be submitted to the TC Chair, Dr. Li‐Chen Ou at the National Taiwan University of Science and Technology. © 2012 Wiley Periodicals, Inc. Col Res Appl, 2012     

《汉英播音主持艺术词典》编纂的可行性浅探

汉英播音主持艺术词典的编纂是建设中国播音学学科的一个重要组成部分。文章从词典编纂性质与意义、理论与架构、方法与范例、前景与展望等4个方面分析了编纂的构想和思路,指出了编纂的必要性和重要意义。     

Instrumentality,aesthetics and symbolism of physical artifacts as triggers of emotion

Building on qualitative data collected from three groups of professionals who assessed the green colour of a public transportation bus, this paper develops a model of the relationship between physical artifacts and emotions. The model suggests that artifacts need to be analysed according to three conceptually distinct aspects: instrumentality, aesthetics and symbolism. These three aspects are suggested to arouse emotion through different mechanisms: a hygiene, a sensory and an associative mechanism. The model opens an arena for extensive future research on the role and influence of physical artifacts in general and on emotions in particular.     

Output-associative RVM regression for dimensional and continuous emotion prediction

Many problems in machine learning and computer vision consist of predicting multi-dimensional output vectors given a specific set of input features. In many of these problems, there exist inherent temporal and spatial dependencies between the output vectors, as well as repeating output patterns and input–output associations, that can provide more robust and accurate predictors when modeled properly. With this intrinsic motivation, we propose a novel Output-Associative Relevance Vector Machine (OA-RVM) regression framework that augments the traditional RVM regression by being able to learn non-linear input and output dependencies. Instead of depending solely on the input patterns, OA-RVM models output covariances within a predefined temporal window, thus capturing past, current and future context. As a result, output patterns manifested in the training data are captured within a formal probabilistic framework, and subsequently used during inference. As a proof of concept, we target the highly challenging problem of dimensional and continuous prediction of emotions, and evaluate the proposed framework by focusing on the case of multiple nonverbal cues, namely facial expressions, shoulder movements and audio cues. We demonstrate the advantages of the proposed OA-RVM regression by performing subject-independent evaluation using the SAL database that constitutes naturalistic conversational interactions. The experimental results show that OA-RVM regression outperforms the traditional RVM and SVM regression approaches in terms of accuracy of the prediction (evaluated using the Root Mean Squared Error) and structure of the prediction (evaluated using the correlation coefficient), generating more accurate and robust prediction models.     

Representation of facial expression categories in continuous arousal–valence space: Feature and correlation

Representation of facial expressions using continuous dimensions has shown to be inherently more expressive and psychologically meaningful than using categorized emotions, and thus has gained increasing attention over recent years. Many sub-problems have arisen in this new field that remain only partially understood. A comparison of the regression performance of different texture and geometric features and the investigation of the correlations between continuous dimensional axes and basic categorized emotions are two of these. This paper presents empirical studies addressing these problems, and it reports results from an evaluation of different methods for detecting spontaneous facial expressions within the arousal–valence (AV) dimensional space. The evaluation compares the performance of texture features (SIFT, Gabor, LBP) against geometric features (FAP-based distances), and the fusion of the two. It also compares the prediction of arousal and valence, obtained using the best fusion method, to the corresponding ground truths. Spatial distribution, shift, similarity, and correlation are considered for the six basic categorized emotions (i.e. anger, disgust, fear, happiness, sadness, surprise). Using the NVIE database, results show that the fusion of LBP and FAP features performs the best. The results from the NVIE and FEEDTUM databases reveal novel findings about the correlations of arousal and valence dimensions to each of six basic emotion categories.