共查询到19条相似文献,搜索用时 78 毫秒
1.
图像的显著区域代表的是图像的主要内容,提取显著区域在图像检索、压缩、识别等方面有重要的应用。基于Itti模型提出一种改进的模型来提取图像的显著区域,即分别提取原始图像的亮度、颜色、方向、形状四个特征,并将这四个特征的多尺度图像特征合并成一幅总的显著图;在图像方向特征提取中采用倒数函数来代替Gabor特征函数,使得方向特征图更加完善。该模型与经典的Itti模型相比较,提取出的显著图更为明显,速度较快且易于实现。 相似文献
2.
《现代电子技术》2018,(10):183-186
针对Itti视觉选择性注意模型不具有子特征图显著图归一化过程中权值随任务改变而改变的问题,借鉴自主发育在视觉选择性注意学习的研究成果,提出一种权值可发育视觉选择性注意模型作为图像特征提取的学习机制。该算法采用三层自组织神经网络和Itti视觉选择性注意模型相结合的决策进行寻优,通过对模型的训练学习获取最优权值更新。这样既可以保证在初期特征提取内容的完整性,又降低了系统对不同任务条件的约束性,提高了模型特征提取能力。利用权值可发育视觉选择性注意模型对图像进行感兴趣区域特征提取实验,结果表明,该方法能够提高特征提取准确性、减少运算时间,获得了良好的动态性能。 相似文献
3.
一种基于视觉显著图的舰船红外图像目标检测方法 总被引:1,自引:0,他引:1
提出了一种基于视觉显著图的红外舰船目标定位方法,即通过改进的Itti模型生成视觉显著图,并基于视觉显著图分割出目标区域,从而实现目标检测。先用小波变换替代Itti模型中的高斯滤波来生成图像多尺度金字塔,然后用center—surround算子提取出多尺度的视觉差异特征,并对生成的视觉特征图进行合成,生成显著图。最后,利用阈值分割方法分割出目标区域,并对原始图像进行标记,从而实现目标检测。实验结果表明,与传统的Otsu阈值分割方法相比,该方法能够准确检测出目标区域。 相似文献
4.
5.
6.
7.
在语音情感识别技术中,由于噪声环境、说话方式和说话人特质原因,会造成实验数据库特征不匹配的情况。从语音学上分析,该问题多存在于跨数据库情感识别实验。训练的声学模型和用于测试的语句样本之间的错位,会使语音情感识别性能剧烈下降。本文据此所研究的选择性注意声学模型能有效探测变化的情感特征。同时,利用时频原子对模型进行改进,使之能提取跨语音数据库中的显著性特征用于情感识别。实验结果表明,利用文章所提方法在跨库情感样本上进行特征提取,再通过典型的分类器,识别性能提高了9个百分点,从而验证了该方法对不同数据库具有更好的鲁棒性。 相似文献
8.
9.
一种新的听觉感知模型 总被引:2,自引:0,他引:2
本文根据心理声学的基本理论和实验资料,提出了一个基于听觉感知特征的听觉模型,与基一生理声学的听觉模型相比较,该模型更能反映人对连续自然语音的感知特性,因而适于对连续语音的识别。 相似文献
10.
本文根据心理声学的基本理论和实验资料,提出了一个基于听觉感知特征的听觉模型,与基于生理声学的听觉模型相比较,该模型更能反映人对连续自然语音的感知特性,因而适于对连续语音的识别。 相似文献
11.
分析了基于自底向上的视觉注意计算模型的感兴趣区域检测方法;它分别提取颜色、灰度、纹理三个特征图像,然后进行线性融合得到综合显著图。而显著目标通常自身灰度相近,但与背景灰度不同,根据这个特性结合灰度概率统计方法对视觉注意计算模型进行改进。实验结果验证了,该模型能够更好的模拟视觉注意的过程,而且计算复杂度较低。 相似文献
12.
Using computational auditory models to predict simultaneous masking data: model comparison 总被引:1,自引:0,他引:1
In order to develop improved remediation techniques for hearing impairment, auditory researchers must gain a greater understanding of the relation between the psychophysics of hearing and the underlying physiology. One approach to studying the auditory system has been to design computational auditory models that predict neurophysiological data such as neural firing rates [15], [1]. To link these physiologically-based models to psychophysics, theoretical bounds on detection performance have been derived using signal detection theory to analyze the simulated data for various psychophysical tasks [20]. Previous efforts, including our own recent work using the Auditory Image Model, have demonstrated the validity of this type of analysis; however, theoretical predictions often continue to exceed experimentally-measured performance [9], [21]. In this paper, we compare predictions of detection performance across several computational auditory models. We also reconcile some of the previously observed discrepancies by incorporating appropriate signal uncertainty into the optimal detector. 相似文献
13.
14.
The human visual system has the ability to rapidly identify and redirect attention to important visual information in high complexity scenes such as the human crowd. Saliency prediction in the human crowd scene is the process using computer vision techniques to imitate the human visual system, predicting which areas in a human crowd scene may attract human attention. However, it is a challenging task to identify which factors may attract human attention due to the high complexity of the human crowd scene. In this work, we propose Multiscale DenseNet — Dilated and Attention (MSDense-DAt), a convolutional neural network (CNN) using self-attention to integrate the result of knowledge-driven gaze in the human visual system to identify salient areas in the human crowd scene. Our method combines various state-of-the-art deep learning architectures to deal with the high complexity in human crowd image, such as multiscale DenseNet for multiscale deep features extraction, self-attention, and dilated convolution. Then the effectiveness of each component in our CNN architecture is evaluated by comparing different components combinations. Finally, the proposed method is further evaluated in different crowd density levels to appraise the effect of crowd density on model performance. 相似文献
15.
16.
17.
深度信息被证明是人类视觉的重要组成部分,然而大部分显著性检测工作侧重于2维图像上的方法,并不能很好地利用深度进行RGB-D图像显著性检测。该文提出一种融合显著深度特征的RGB-D图像显著目标检测方法,提取基于颜色和深度显著图的综合特征,根据构图先验和背景先验的方法进行显著目标检测。首先,对原始深度图进行预处理:使用背景顶点区域、构图交点和紧密度处理深度图,多角度融合形成深度显著图,并作为显著深度特征,结合颜色特征形成综合特征;其次,从前景角度,将综合特征通过边连接权重构造关联矩阵,根据构图先验,假设多层中心矩形为前景种子,通过流形排序方法计算出RGB-D图像的前景显著图;从背景角度,根据背景先验以及边界连通性计算出背景显著图;最后,将前景显著图和背景显著图进行融合并优化得到最终显著图。实验采用RGB-D1000数据集进行显著性检测,并与4种不同的方法进行对比,所提方法的显著性检测结果更接近人工标定结果,PR(查准率-查全率)曲线显示在相同召回率下准确率高于其他方法。 相似文献
18.
On the computational model of a kind of deconvolution problem 总被引:2,自引:0,他引:2
It is known that discretization of a continuous deconvolution problem can alleviate the ill-posedness of the problem. The currently used circulant matrix model, however, does not play such a role. Moreover, the approximation of deconvolution problems by circulant matrix model is rational only if the size of the kernel function is very small. We propose an aperiodic model of deconvolution. For discrete and finite deconvolution problems the new model is an exact one. In the general case, the new model can lead to a nonsingular system of equations that has a lower condition number than the circulant one, and the related computations in the deconvolution can be done efficiently by means of the DFT technique, as in the ease for circulant matrices. The rationality of the new model holds without regard to the size of the kernel and the image. The use of the aperiodic model is illustrated by gradient-based algorithms. 相似文献