Lightweight network architecture using difference saliency maps for facial action unit detection

Facial action unit (AU) detection has been applied in a wild range of fields, and has attracted great attention over the last decades. Most existing methods employ the predefined regions of interest with same number and range for all samples. However, we find that the flexibility of predefined regions of interest is finite, as the occurrence of different AUs may not be simultaneous and their ranges change with intensity changes. In addition, many AU detection works try to independently design feature extraction modules and classifiers for each AU, which is of high computation cost and ignores the dependency among different AUs. In view of the limited flexibility of predefined regions of interest, we propose difference saliency maps that do not depend on facial landmarks. They are the spatial pixel-wise attentions, where each element represents the importance of the corresponding pixel on the entire image. Therefore, all the regions of interest can be irregular. In addition, in order to solve the problem of high computation cost, we combine group convolution with skip connection to propose a lightweight network that is more suitable for AU detection. All AUs share features and there is only one classifier, so the computation cost and the number of parameters are greatly reduced. In particular, the difference saliency maps and the global feature maps are combined to obtain the regional enhancement features. To maximize the enhancement effect, the down-sampled difference saliency maps are added to multiple blocks of the lightweight network. The enhanced global features are directly sent to the classifier for AU detection. By changing the number of neurons in the classifier, our framework can easily adapt to different datasets. Extensive experimental results show that the proposed framework soundly outperforms the classic deep learning method when evaluated on the DISFA+ and CK+ datasets. After adding the difference saliency maps, the detection result is better than the state-of-the-art AU detection methods. Further experiments demonstrate that our network is more efficient in using parameters, computation complexity and inference time.

     

融合双目多维感知特征的立体视频显著性检测

目的 立体视频能提供身临其境的逼真感而越来越受到人们的喜爱,而视觉显著性检测可以自动预测、定位和挖掘重要视觉信息,可以帮助机器对海量多媒体信息进行有效筛选。为了提高立体视频中的显著区域检测性能,提出了一种融合双目多维感知特性的立体视频显著性检测模型。方法 从立体视频的空域、深度以及时域3个不同维度出发进行显著性计算。首先,基于图像的空间特征利用贝叶斯模型计算2D图像显著图;接着,根据双目感知特征获取立体视频图像的深度显著图;然后,利用Lucas-Kanade光流法计算帧间局部区域的运动特征,获取时域显著图;最后,将3种不同维度的显著图采用一种基于全局-区域差异度大小的融合方法进行相互融合,获得最终的立体视频显著区域分布模型。结果 在不同类型的立体视频序列中的实验结果表明,本文模型获得了80%的准确率和72%的召回率,且保持了相对较低的计算复杂度,优于现有的显著性检测模型。结论 本文的显著性检测模型能有效地获取立体视频中的显著区域,可应用于立体视频/图像编码、立体视频/图像质量评价等领域。     

超像素条件随机场下的RGB-D视频显著性检测

目的 视觉显著性在众多视觉驱动的应用中具有重要作用,这些应用领域出现了从2维视觉到3维视觉的转换,从而基于RGB-D数据的显著性模型引起了广泛关注。与2维图像的显著性不同,RGB-D显著性包含了许多不同模态的线索。多模态线索之间存在互补和竞争关系,如何有效地利用和融合这些线索仍是一个挑战。传统的融合模型很难充分利用多模态线索之间的优势,因此研究了RGB-D显著性形成过程中多模态线索融合的问题。方法 提出了一种基于超像素下条件随机场的RGB-D显著性检测模型。提取不同模态的显著性线索,包括平面线索、深度线索和运动线索等。以超像素为单位建立条件随机场模型,联合多模态线索的影响和图像邻域显著值平滑约束,设计了一个全局能量函数作为模型的优化目标,刻画了多模态线索之间的相互作用机制。其中,多模态线索在能量函数中的权重因子由卷积神经网络学习得到。结果 实验在两个公开的RGB-D视频显著性数据集上与6种显著性检测方法进行了比较,所提模型在所有相关数据集和评价指标上都优于当前最先进的模型。相比于第2高的指标,所提模型的AUC（area under curve）,sAUC（shuffled AUC）,SIM（similarity）,PCC（Pearson correlation coefficient）和NSS（normalized scanpath saliency）指标在IRCCyN数据集上分别提升了2.3%,2.3%,18.9%,21.6%和56.2%;在DML-iTrack-3D数据集上分别提升了2.0%,1.4%,29.1%,10.6%,23.3%。此外还进行了模型内部的比较,验证了所提融合方法优于其他传统融合方法。结论 本文提出的RGB-D显著性检测模型中的条件随机场和卷积神经网络充分利用了不同模态线索的优势,将它们有效融合,提升了显著性检测模型的性能,能在视觉驱动的应用领域发挥一定作用。     

应用奇异值分解的海上场景显著性检测

提出一种应用奇异值分解的海上场景显著性检测方法。提取海上场景图像中颜色和亮度各通道特征,并对各其分别进行奇异值分解,根据设定的阈值,选择各特征的典型分量。各特征的粗显著图定义为各特征和其典型分量的差。为进一步去除海杂波等干扰,在粗显著图中,计算其空间域全局显著性,以此形成显著性图。得到的颜色通道和亮度通道显著图通过线性合并为总显著图。利用海上场景图像进行了实验,结果表明提出方法的有效性。     

多尺度卷积神经网络显著物体检测

目的 传统显著性检测模型大多利用手工选择的中低层特征和先验信息进行物体检测,其准确率和召回率较低,随着深度卷积神经网络的兴起,显著性检测得以快速发展。然而,现有显著性方法仍存在共性缺点,难以在复杂图像中均匀地突显整个物体的明确边界和内部区域,主要原因是缺乏足够且丰富的特征用于检测。方法 在VGG（visual geometry group）模型的基础上进行改进,去掉最后的全连接层,采用跳层连接的方式用于像素级别的显著性预测,可以有效结合来自卷积神经网络不同卷积层的多尺度信息。此外,它能够在数据驱动的框架中结合高级语义信息和低层细节信息。为了有效地保留物体边界和内部区域的统一,采用全连接的条件随机场（conditional random field,CRF）模型对得到的显著性特征图进行调整。结果 本文在6个广泛使用的公开数据集DUT-OMRON（Dalian University of Technology and OMRON Corporation）、ECSSD（extended complex scene saliency dataset）、SED2（segmentation evalution database 2）、HKU、PASCAL-S和SOD（salient objects dataset）上进行了测试,并就准确率—召回率（precision-recall,PR）曲线、F测度值（F-measure）、最大F测度值、加权F测度值和均方误差（mean absolute error,MAE）等性能评估指标与14种最先进且具有代表性的方法进行比较。结果显示,本文方法在6个数据集上的F测度值分别为0.696、0.876、0.797、0.868、0.772和0.785;最大F测度值分别为0.747、0.899、0.859、0.889、0.814和0.833;加权F测度值分别为0.656、0.854、0.772、0.844、0.732和0.762;MAE值分别为0.074、0.061、0.093、0.049、0.099和0.124。无论是前景和背景颜色相似的图像集,还是多物体的复杂图像集,本文方法的各项性能均接近最新研究成果,且优于大多数具有代表性的方法。结论 本文方法对各种场景的图像显著性检测都具有较强的鲁棒性,同时可以使显著性物体的边界和内部区域更均匀,检测结果更准确。     

Images as sets of locally weighted features

This paper presents a generic framework in which images are modelled as order-less sets of weighted visual features. Each visual feature is associated with a weight factor that may inform its relevance. This framework can be applied to various bag-of-features approaches such as the bag-of-visual-word or the Fisher kernel representations. We suggest that if dense sampling is used, different schemes to weight local features can be evaluated, leading to results that are often better than the combination of multiple sampling schemes, at a much lower computational cost, because the features are extracted only once. This allows our framework to be a test-bed for saliency estimation methods in image categorisation tasks. We explored two main possibilities for the estimation of local feature relevance. The first one is based on the use of saliency maps obtained from human feedback, either by gaze tracking or by mouse clicks. The method is able to profit from such maps, leading to a significant improvement in categorisation performance. The second possibility is based on automatic saliency estimation methods, including Itti & Koch’s method and SIFT’s DoG. We evaluated the proposed framework and saliency estimation methods using an in house dataset and the PASCAL VOC 2008/2007 dataset, showing that some of the saliency estimation methods lead to a significant performance improvement in comparison to the standard unweighted representation.     

Pedestrian object detection with fusion of visual attention mechanism and semantic computation

In response to the problem that the primary visual features are difficult to effectively address pedestrian detection in complex scenes, we present a method to improve pedestrian detection using a visual attention mechanism with semantic computation. After determining a saliency map with a visual attention mechanism, we can calculate saliency maps for human skin and the human head-shoulders. Using a Laplacian pyramid, the static visual attention model is established to obtain a total saliency map and then complete pedestrian detection. Experimental results demonstrate that the proposed method achieves state-of-the-art performance on the INRIA dataset with 92.78% pedestrian detection accuracy at a very competitive time cost.

     

不规则像素簇显著性检测算法

目的 显著性检测领域的研究重点和难点是检测具有复杂结构信息的显著物体。传统的基于图像块的检测算法,主要根据相对规则的图像块进行检测,计算过程中不能充分利用图像不规则的结构和纹理的信息,对算法精度产生影响。针对上述问题,本文提出一种基于不规则像素簇的显著性检测算法。方法 根据像素点的颜色信息量化颜色空间,同时寻找图像的颜色中心,将每个像素的颜色替代为最近的颜色中心的颜色。然后根据相同颜色标签的连通域形成不规则像素簇,并以连通域的中心为该簇的位置中心,以该连通域对应颜色中心的颜色为该簇整体的颜色。通过像素簇的全局对比度得到对比度先验图,利用目标粗定位法估计显著目标的中心,计算图像的中心先验图。然后将对比度先验图与中心先验图结合得到初始显著图。为了使显著图更加均匀地突出显著目标,利用图模型及形态学变化改善初始显著图效果。结果 将本文算法与5种公认表现最好的算法进行对比,并通过5组图像进行验证,采用客观评价指标精确率—召回率（precision-recall,PR）曲线以及精确率和召回率的调和平均数F-measure进行评价,结果表明本文算法在PR曲线上较其他算法表现良好,在F-measure方面相比其他5种算法均有00.3的提升,且有更佳的视觉效果。结论 本文通过更合理地对像素簇进行划分,并对目标物体进行粗定位,更好地考虑了图像的结构和纹理特征,在显著性检测中有较好的检测效果,普适性强。     

Saliency detection via image sparse representation and color features combination

Saliency detection is a technique to analyze image surroundings to extract relevant regions from the background. In this paper, we propose a simple and effective saliency detection method based on image sparse representation and color features combination. First, the input image is segmented into non-overlapping super-pixels, so as to perform the saliency detection at the region level to reduce computational complexity. Then, a background optimization selection scheme is used to construct an appropriate background template. Based on this, a primary saliency map is obtained by using image sparse representation. Next, through the linear combination of color coefficients we generate an improved saliency map with more prominent salient regions. Finally, the two saliency maps are integrated within Bayesian framework to obtain the final saliency map. Experimental results show that the proposed method has desirable detection performance in terms of detection accuracy and running time.

     

特征融合与S-D概率矫正的RGB-D显著检测

目的 许多先前的显著目标检测工作都是集中在2D的图像上,并不能适用于RGB-D图像的显著性检测。本文同时提取颜色特征以及深度特征,提出了一种基于特征融合和S-D概率矫正的RGB-D显著性检测方法,使得颜色特征和深度特征相互补充。方法 首先,以RGB图像的4个边界为背景询问节点,使用特征融合的Manifold Ranking输出RGB图像的显著图;其次,依据RGB图像的显著图和深度特征计算S-D矫正概率;再次,计算深度图的显著图并依据S-D矫正概率对该显著图进行S-D概率矫正;最后,对矫正后的显著图提取前景询问节点再次使用特征融合的Manifold Ranking方法进行显著优化,得到最终的显著图。结果 利用本文RGB-D显著性检测方法对RGBD数据集上的1 000幅图像进行了显著性检测,并与6种不同的方法进行对比,本文方法的显著性检测结果更接近人工标定结果。Precision-Recall曲线（PR曲线）显示在相同召回率下本文方法的准确率较其中5种方法高,且处理单幅图像的时间为2.150 s,与其他算法相比也有一定优势。结论 本文方法能较准确地对RGB-D图像进行显著性检测。     

多先验特征与综合对比度的图像显著性检测

目的 图像的显著性检测在计算机视觉中应用非常广泛,现有的方法通常在复杂背景区域下表现不佳,由于显著性检测的低层特征并不可靠,同时单一的特征也很难得到高质量的显著图。提出了一种通过增加特征的多样性来实现显著性检测的方法。方法 在高层先验知识的基础上,对背景先验特征和中心先验特征重新进行了定义,并考虑人眼视觉一般会对暖色调更为关注,从而加入颜色先验。另外在图像低层特征上使用目前较为流行的全局对比度和局部对比度特征,在特征融合时针对不同情况分别采取线性和非线性的一种新的融合策略,得到高质量的显著图。结果 在MSRA-1000和DUT-OMRON两个公开数据库进行对比验证,实验结果表明,基于多先验特征与综合对比度的图像显著性检测算法具有较高的查准率、召回率和F-measure值,相较于RBD算法均提高了1.5%以上,综合性能均优于目前的10种主流算法。结论 相较于基于低层特征和单一先验特征的算法,本文算法充分利用了图像信息,能在突出全局对比度的同时也保留较多的局部信息,达到均匀突出显著性区域的效果,有效地抑制复杂的背景区域,得到更加符合视觉感知的显著图。     

Saliency detection based on self-adaptive multiple feature fusion for remote sensing images

ABSTRACT

Saliency detection has been revealed an effective and reliable approach to extract the region of interest (ROI) in remote sensing images. However, most existing saliency detection methods employing multiple saliency cues ignore the intrinsic relationship between different cues and do not distinguish the diverse contributions of different cues to the final saliency map. In this paper, we propose a novel self-adaptively multiple feature fusion model for saliency detection in remote sensing images to take advantage of this relationship to improve the accuracy of ROI extraction. First, we take multiple feature channels, namely colour, intensity, texture and global contrast into consideration to produce primary feature maps. Particularly, we design a novel method based on dual-tree complex wavelet transform for remote sensing images to generate texture feature pyramids. Then, we introduce a novel self-adaptive multiple feature fusion method based on low-rank matrix recovery, in which the significances of feature maps are ranked by the low rank constraint recovery, and subsequently multiple features’ contributions are allocated adaptively to produce the final saliency map. Experimental results demonstrate that our proposal outperforms the state-of-the-art methods.     

利用视觉显著性的图像分割方法

提出一种利用视觉显著性对图像进行分割的方法。首先提取图像的底层视觉特征,从局部显著性、全局显著性和稀少性3个方面计算各特征图像中各像素的视觉显著性,得到各特征显著图;对各特征显著图进行综合,生成最终的综合显著图。然后对综合显著图进行阈值分割,得到二值图像,将二值图像与原始图像叠加,将前景和背景分离,得到图像分割结果。在多幅自然图像上进行实验验证,并给出相应的实验结果和分析。实验结果表明,该方法正确有效,具有和人类视觉特性相符合的分割效果。     

融合深度模型和传统模型的显著性检测

目的 显著性检测是图像和视觉领域一个基础问题,传统模型对于显著性物体的边界保留较好,但是对显著性目标的自信度不够高,召回率低,而深度学习模型对于显著性物体的自信度高,但是其结果边界粗糙,准确率较低。针对这两种模型各自的优缺点,提出一种显著性模型以综合利用两种方法的优点并抑制各自的不足。方法 首先改进最新的密集卷积网络,训练了一个基于该网络的全卷积网络（FCN）显著性模型,同时选取一个现有的基于超像素的显著性回归模型,在得到两种模型的显著性结果图后,提出一种融合算法,融合两种方法的结果以得到最终优化结果,该算法通过显著性结果Hadamard积和像素间显著性值的一对一非线性映射,将FCN结果与传统模型的结果相融合。结果 实验在4个数据集上与最新的10种方法进行了比较,在HKU-IS数据集中,相比于性能第2的模型,F值提高了2.6%;在MSRA数据集中,相比于性能第2的模型,F值提高了2.2%,MAE降低了5.6%;在DUT-OMRON数据集中,相比于性能第2的模型,F值提高了5.6%,MAE降低了17.4%。同时也在MSRA数据集中进行了对比实验以验证融合算法的有效性,对比实验结果表明提出的融合算法改善了显著性检测的效果。结论 本文所提出的显著性模型,综合了传统模型和深度学习模型的优点,使显著性检测结果更加准确。     

自动特征选择和加权的图像显著区域检测

视觉显著性度量是图像显著区域提取中的一个关键问题,现有的方法主要根据图像的底层视觉特征,构造相应的显著图。不同的特征对视觉显著性的贡献是不同的,为此提出一种能够自动进行特征选择和加权的图像显著区域检测方法。提取图像的亮度、颜色和方向等特征,构造相应的特征显著图。提出一种新的特征融合策略,动态计算各特征显著图的权值,整合得到最终的显著图,检测出图像中的显著区域。在多幅自然图像上进行实验,实验结果表明,该方法在运算速度和检测效果方面都取得了不错的效果。     

稀疏深度特征对传统显著性检测的优化

目的 显著性目标检测算法主要分为基于低级特征的传统方法和基于深度学习的新方法,传统方法难以捕获对象的高级语义信息,基于深度学习的新方法能捕获高级语义信息却忽略了边缘特征。为了充分发挥两种方法的优势,基于将二者结合的思路,本文利用稀疏能使得显著性对象指向性凝聚的优势,提出了一种基于稀疏自编码和显著性结果优化的方法。方法 对VGG（visual geometry group）网络第4个池化层的特征图进行稀疏自编码处理,得到5张稀疏显著性特征图,再与传统方法得到的显著图一起输入卷积神经网络进行显著性结果优化。结果 使用DRFI（discriminative regional feature integration）、HDCT（high dimensional color transform）、RRWR（regularized random walks ranking）和CGVS（contour-guided visual search）等传统方法在DUT-OMRON、ECSSD、HKU-IS和MSRA等公开数据集上进行实验,表明本文算法有效改善了显著性对象的F值和MAE（mean absolute error）值。在F值提高方面,优化后的DRFI方法提升最高,在HKU-IS数据集上提高了24.53%。在MAE值降低方面,CGVS方法降低最少,在ECSSD数据集上降低了12.78%,降低最多的接近50%。而且本模型结构简单,参数少,计算效率高,训练时间约5 h,图像的平均测试时间约为3 s,有很强的实际应用性。结论 本文提出了一种显著性结果优化算法,实验结果表明算法有效改善了显著性对象F值和MAE值,在对显著性对象检测要求越来越准确的对象识别等任务中有较好的适应性和应用性前景。     

Color boosted visual saliency detection and its application to image classification

For many applications in graphics, design and human computer interaction, it is essential to reliably estimate the visual saliency of images. In this paper, we propose a visual saliency detection method that combines the respective merits of color saliency boosting and global region based contrast schemes to achieve more accurate saliency maps. Our method is compared with existing saliency detection methods when evaluated using four public available datasets. Experimental results show that our method consistently outperformed current state-of-the-art methods on predicting human fixations. We also demonstrate how the extracted saliency map can be used for image classification.     

A Two-Stage Bayesian Integration Framework for Salient Object Detection on Light Field

Unique visual features of 4D light field data have been shown to affect detection of salient objects. Nevertheless, only a few studies explore it yet. In this study, several helpful visual features extracted from light field data are fused in a two-stage Bayesian integration framework for salient object detection. First, background weighted color contrast is computed in high dimensional color space, which is more distinctive to identify object of interest. Second, focusness map of foreground slice is estimated. Then, it is combined with the color contrast results via first-stage Bayesian fusion. Third, background weighted depth contrast is computed. Depth contrast has been proved to be an extremely useful cue for salient object detection and complementary to color contrast. Finally, in the second-stage Bayesian fusion step, the depth-induced contrast saliency is further fused with the first-stage saliency fusion results to get the final saliency map. Experiments of comparing with eight existing state-of-the-art methods on light field benchmark datasets show that the proposed method can handle challenging scenarios such as cluttered background, and achieves the most visually acceptable salient object detection results.     

融合多特征与先验信息的显著性目标检测

目的 图像的显著性目标检测是计算机视觉领域的重要研究课题。针对现有显著性目标检测结果存在的纹理细节刻画不明显和边缘轮廓显示不完整的问题,提出一种融合多特征与先验信息的显著性目标检测方法,该方法能够高效而全面地获取图像中的显著性区域。方法 首先,提取图像感兴趣的点集,计算全局对比度图,利用贝叶斯方法融合凸包和全局对比度图获得对比度特征图。通过多尺度下的颜色直方图得到颜色空间图,根据信息熵定理计算最小信息熵,并将该尺度下的颜色空间图作为颜色特征图。通过反锐化掩模方法提高图像清晰度,利用局部二值算子（LBP）获得纹理特征图。然后,通过图形正则化（GR）和流行排序（MR）算法得到中心先验图和边缘先验图。最后,利用元胞自动机融合对比度特征图、颜色特征图、纹理特征图、中心先验图和边缘先验图获得初级显著图,再通过快速引导滤波器优化处理得到最终显著图。结果 在2个公开的数据集MSRA10K和ECSSD上验证本文算法并与12种具有开源代码的流行算法进行比较,实验结果表明,本文算法在准确率-召回率（PR）曲线、受试者工作特征（ROC）曲线、综合评价指标（F-measure）、平均绝对误差（MAE）和结构化度量指标（S-measure）等方面有显著提升,整体性能优于对比算法。结论 本文算法充分利用了图像的对比度特征、颜色特征、纹理特征,采用中心先验和边缘先验算法,在全面提取显著性区域的同时,能够较好地保留图像的纹理信息和细节信息,使得边缘轮廓更加完整,满足人眼的层次要求和细节要求,并具有一定的适用性。