Familiarity based unified visual attention model for fast and robust object recognition

Even though visual attention models using bottom-up saliency can speed up object recognition by predicting object locations, in the presence of multiple salient objects, saliency alone cannot discern target objects from the clutter in a scene. Using a metric named familiarity, we propose a top-down method for guiding attention towards target objects, in addition to bottom-up saliency. To demonstrate the effectiveness of familiarity, the unified visual attention model (UVAM) which combines top-down familiarity and bottom-up saliency is applied to SIFT based object recognition. The UVAM is tested on 3600 artificially generated images containing COIL-100 objects with varying amounts of clutter, and on 126 images of real scenes. The recognition times are reduced by 2.7× and 2×, respectively, with no reduction in recognition accuracy, demonstrating the effectiveness and robustness of the familiarity based UVAM.     

视觉选择性注意模型化计算中的特征整合权值估计与图像显著性区域提取

针对视觉选择性注意模型化计算过程中不同特征在整合阶段的权值判定,提出一种基于特征图分布的权值估计方法,并在静态图像显著性区域提取中取得了令人满意的应用效果。首先提取原始图像的颜色、方向和强度特征图像,然后计算各个特征图的广义高斯分布参数与方差,进而给出一种特征图权值估计算法,最后通过对特征图的加权整合与归一化实现对原始图像的显著性区域提取。实验结果表明,通过此方法计算的权值对特征进行加权调制所提取的显著性区域的效果更加符合人眼的观测结果。     

一种面向视觉注意区域检测的运动分割方法*

结合视觉显著区检测的特点,本文提出一种面向视觉注意区域检测的运动分割方法。该方法用一种层次聚类方法将特征点的运动轨迹进行聚类。首先用中值偏移算法扩大了不同类型运动之间特征向量的差距,同时缩小了相同运动类型的差别。继而,用一种无监督聚类算法,将不同类型的运动进行分割,同时自动获得运动分类数。最后利用运动分割结果,提出一种结合空间和颜色采样的运动显著区域生成方法。与以往方法相比,该方法能够将不同类型的运动自动进行分割,生成的视觉注意区域更为准确,而且稳定性大幅提高。实验结果证明了该方法的有效性和稳定性。     

基于Bayes与多层Bayes估计的WSN链路选择算法

无线传感器网络（WSN）节点能量有限,采用传统的链路选择的方法（经验法）进行链路选择,需要发送大量的数据包作为测试样本,这在WSN中是不合适的。设计了两种基于Bayes估计与一种基于多层Bayes估计的WSN链路选择算法,分别记为BLSP-B1、BLSP-B2、BLSP-HE。仿真实验发现,在小样本的条件下,BLSP-B1、BLSP-B2、BLSP-HE选择高质量的链路的概率比经验法要高出10%～20%,其中BLSP-HE算法最稳健,性能较好。     

基于视觉显著熵与Object Bank特征的图像记忆性模型

为了提高图像的记忆性预测能力,提出了一种基于视觉显著熵与改进的Object Bank特征的图像记忆性自动预测方法。该方法改进了传统的Object Bank特征,提取图像的视觉显著熵特征,利用支持向量回归机(SVR)训练得到图像的记忆性预测模型。实验结果表明,在预测准确性方面,所提方法比现有的方法的相关系数高出3个百分点。所提出的模型可以应用于图像的记忆性预测、图像检索排序、广告评价分析等方向。     

The degree of human visual attention in the visual search

Human beings can obtain visual information in parallel through the retina, but they cannot pay attention to all the information at the same time. In psychological studies, the human characteristics of visual attention have often been investigated by analyzing the characteristics of the visual search task. Previous studies suggested that the information features of the visual search task are processed in parallel at early stages of processing. However, the authors consider that these features are not processed completely in parallel, and have a reciprocal action to each other. In order to clarify the reciprocal action of the features in a visual search and the continuity of visual attention, the characteristics of reaction times were measured with changing forms of visual stimuli. The experimental results suggested that the reaction time changed when the features of the visual stimuli in the visual search task changed. This means that the features are affected by each other. Furthermore, continuity of reciprocal action is also suggested, and the degree of visual attention is decided by this continuity. The results provided significant basic data to support our proposed mathematical model of visual attention. This work was presented, in part, at the Fourth International Symposium on Artificial Life and Robotics, Oita, Japan, January 19–22, 1999     

Articulatory feature recognition using dynamic Bayesian networks

We describe a dynamic Bayesian network for articulatory feature recognition. The model is intended to be a component of a speech recognizer that avoids the problems of conventional “beads-on-a-string” phoneme-based models. We demonstrate that the model gives superior recognition of articulatory features from the speech signal compared with a state-of-the-art neural network system. We also introduce a training algorithm that offers two major advances: it does not require time-aligned feature labels and it allows the model to learn a set of asynchronous feature changes in a data-driven manner.     

机械臂视觉伺服系统中的高精度实时特征提取

以机械臂视觉伺服系统为背景,立体视觉所需要的特征点为目标,提出一种基于颜色块的实时特征提取方法．特征点的定位分为粗定位和精定位两个过程,粗定位用来跟踪目标颜色块,精定位在粗定位的基础上,切出小图像并提取目标轮廓,利用犓犔变换对轮廓点进行分类,并通过最小二乘法对边缘点进行拟合．根据直线方程计算顶点的亚像素精度坐标,为下一步的目标位姿测量提供基础．实验结果验证了该方法的有效性．     

Bayesian shape model for facial feature extraction and recognition

A facial feature extraction algorithm using the Bayesian shape model (BSM) is proposed in this paper. A full-face model consisting of the contour points and the control points is designed to describe the face patch, using which the warping/normalization of the extracted face patch can be performed efficiently. First, the BSM is utilized to match and extract the contour points of a face. In BSM, the prototype of the face contour can be adjusted adaptively according to its prior distribution. Moreover, an affine invariant internal energy term is introduced to describe the local shape deformations between the prototype contour in the shape domain and the deformable contour in the image domain. Thus, both global and local shape deformations can be tolerated. Then, the control points are estimated from the matching result of the contour points based on the statistics of the full-face model. Finally, the face patch is extracted and normalized using the piece-wise affine triangle warping algorithm. Experimental results based on real facial feature extraction demonstrate that the proposed BSM facial feature extraction algorithm is more accurate and effective as compared to that of the active shape model (ASM).     

A Bayesian approach for object classification based on clusters of SIFT local features

Several methods have been presented in the literature that successfully used SIFT features for object identification, as they are reasonably invariant to translation, rotation, scale, illumination and partial occlusion. However, they have poor performance for classification tasks. In this work, SIFT features are used to solve object class recognition problems in images using a two-step process. In its first step, the proposed method performs clustering on the extracted features in order to characterize the appearance of the different classes. Then, in the classification step, it uses a three layer Bayesian network for object class recognition. Experiments show quantitatively that clusters of SIFT features are suitable to represent classes of objects. The main contributions of this paper are the introduction of a Bayesian network approach in the classification step to improve performance in an object class recognition task, and a detailed experimentation that shows robustness to changes in illumination, scale, rotation and partial occlusion.     

基于腿部三角特征的贝叶斯步态识别方法

提出了一种基于步态序列中腿部三角特征的步态表示方法,在这种特征上用改进的朴素贝叶斯分类方法进行步态识别。选取步幅最大、最小两种情况下的姿态作为关键帧,用三角型模拟其腿部特征,提取三角型模型参数作为步态特征,识别时先分别用KNN和一种改进的N-best取得属性值在训练数据中的对应数值,然后用贝叶斯分类方法识别。在NLPR数据库上使用留一校验方法进行算法验证,实验证明该方法简单快速,而且取得了比较理想的识别效果。     

Model-based approach for high-dimensional non-Gaussian visual data clustering and feature weighting

Clustering is the task of classifying patterns or observations into clusters or groups. Generally, clustering in high-dimensional feature spaces has a lot of complications such as: the unidentified or unknown data shape which is typically non-Gaussian and follows different distributions; the unknown number of clusters in the case of unsupervised learning; and the existence of noisy, redundant, or uninformative features which normally compromise modeling capabilities and speed. Therefore, high-dimensional data clustering has been a subject of extensive research in data mining, pattern recognition, image processing, computer vision, and other areas for several decades. However, most of existing researches tackle one or two problems at a time which is unrealistic because all problems are connected and should be tackled simultaneously. Thus, in this paper, we propose two novel inference frameworks for unsupervised non-Gaussian feature selection, in the context of finite asymmetric generalized Gaussian (AGG) mixture-based clustering. The choice of the AGG distribution is mainly due to its ability not only to approximate a large class of statistical distributions (e.g. impulsive, Laplacian, Gaussian and uniform distributions) but also to include the asymmetry. In addition, the two frameworks simultaneously perform model parameters estimation as well as model complexity (i.e., both model and feature selection) determination in the same step. This was done by incorporating a minimum message length (MML) penalty in the model learning step and by fading out the redundant densities in the mixture using the rival penalized EM (RPEM) algorithm, for first and second frameworks, respectively. Furthermore, for both algorithms, we tackle the problem of noisy and uninformative features by determining a set of relevant features for each data cluster. The efficiencies of the proposed algorithms are validated by applying them to real challenging problems namely action and facial expression recognition.     

基于稀疏贝叶斯学习的个人信用评估

针对传统信用评估方法分类精度低、特征可解释性差等问题,提出了一种使用稀疏贝叶斯学习方法来进行个人信用评估的模型(SBLCredit)。SBLCredit充分利用稀疏贝叶斯学习的优势,在添加的特征权重的先验知识的情况下进行求解,使得特征权重尽量稀疏,以此实现个人信用评估和特征选择。在德国和澳大利亚真实信用数据集上,SBLCredit方法的分类精度比传统的K近邻、朴素贝叶斯、决策树和支持向量机平均提高了4.52%,6.40%,6.26%和2.27%。实验结果表明,SBLCredit分类精度高,选择的特征少,是一种有效的个人信用评估方法。     

基于支持向量机特征选择的贝叶斯网结点序

目前较常采用搜索打分方法进行贝叶斯网络结构学习,该方法需要首先依据参与者的经验来确定网络的结点顺序,主观性较强,限制了它的实际应用。基于支持向量机特征选择的方法,可以按照各个结点对叶结点的影响能力进行排序,从而直接从数据中通过学习得出结点顺序,避免了人为因素的影响。实验结果验证了该方法的有效性。     

Bayes网络学习的MCMC方法

基于Bayes统计理论, 提出了一种从数据样本中学习Bayes网络的Markov链Monte Carlo(MCMC)方法. 首先通过先验概率和数据样本的结合得到未归一化的后验概率, 然后使用此后验概率指导随机搜索算法寻找“好”的网络结构模型. 通过对Alarm网络的学习表明了本算法具有较好的性能.     

Efficient Discriminant Viewpoint Selection for Active Bayesian Recognition

This paper presents a novel viewpoint selection criterion for active object recognition and pose estimation whose key advantage resides in its low computational cost with respect to current popular approaches in the literature. The proposed observation selection criterion associates high utility with observations that predictably facilitate distinction between pairs of competing hypotheses by a Bayesian classifier. Rigorous experimentation of the proposed approach was conducted on two case studies, involving synthetic and real data, respectively. The results show the proposed algorithm to perform better than a random navigation strategy in terms of the amount of data required for recognition while being much faster than a strategy based on mutual information, without compromising accuracy.     

基于发音特征的音/视频双流语音识别模型

构建了一种基于发音特征的音/视频双流动态贝叶斯网络(dynamic Bayesian network, DBN)语音识别模型,定义了各节点的条件概率关系,以及发音特征之间的异步约束关系,最后在音/视频连接数字语音数据库上进行了语音识别实验,并与音频单流、视频单流DBN模型比较了在不同信噪比情况下的识别效果。结果表明,在低信噪比情况下,基于发音特征的音/视频双流语音识别模型表现出最好的识别性能,而且随着噪声的增加,其识别率下降的趋势比较平缓,表明该模型对噪声具有很强的鲁棒性,更适用于低信噪比环境下的语音识别     

一般贝叶斯网络分类器及其学习算法

贝叶斯网络（BN）应用于分类应用时对目标变量预测有直接贡献的局部模型称作一般贝叶斯网络分类器（GBNC）。推导GBNC的传统途径是先学习完整的BN,而现有推导BN结构的算法限制了应用规模。为了避免学习全局BN,提出仅执行局部搜索的结构学习算法IPC-GBNC,它以目标变量节点为中心执行广度优先搜索,且将搜索深度控制在不超过2层。理论上可证明算法IPC-GBNC是正确的,而基于仿真和真实数据的实验进一步验证了其学习效果和效率的优势：（1）可输出和执行全局搜索的PC算法相同甚至更高质量的结构;（2）较全局搜索消耗少得多的计算量;（3）同时实现了降维（类似决策树学习算法）。相比于绝大多数经典分类器,GBNC的分类性能相当,但兼具直观、紧凑表达和强大推理的能力（且支持不完整观测值）。