Visual processing of contours and textures

这个申请是为了继续我对轮廓和纹理的视觉处理的研究。能够处理轮廓的形状，如梨的弯曲轮廓，以及纹理，如折叠地毯的表面，是我们在视觉世界中识别物体和表面形状的能力的重要组成部分。我的实验涉及人类测试对象观看大量在显示器上显示的专门设计的计算机生成的测试模式，并对它们做出判断。我目前使用的特殊实验方法是“视觉后效”。长时间观看某一轮廓/纹理形状，称为“适应”，会导致其他轮廓/纹理形状出现扭曲。通过测量扭曲程度，我们可以了解到很多关于大脑如何处理轮廓/纹理形状的信息。例如，如果取一个简单的弯曲轮廓并适应它，那么稍微弯曲的轮廓看起来比其他轮廓更弯曲。这种效应被称为“曲率后效”，它发生的事实告诉我们，大脑中有专门用于观察曲线的神经元。因此，曲率后效可以用来揭示大脑如何看待曲线的细节。举一个例子来说明它是如何做到这一点的，我的实验室已经表明，如果适应和测试曲线是不同的颜色，曲率后效就会大大减少（即有更少的失真），相比之下，如果它们是相同的颜色。这告诉我们，大脑中检测曲线的神经元会对颜色进行“调谐”，换句话说，一些神经元会对红色曲线、绿色曲线、蓝色曲线等做出反应。因此，结合其他实验，我的研究可以更好地理解大脑是如何看到轮廓和纹理的形状的。

This application is for funds to continue my research into the visual processing of contours and textures. Being able to process the shapes of contours, such as a curved outline of a pear, as well as textures, such as the surface of a folded carpet, is an essential part of our ability to recognize objects and the shapes of surfaces in the visual world. My experiments involve human test subjects viewing large numbers of specially-designed computer-generated test patterns presented on a display monitor and making judgements about them. The particular experimental method I current employ is the 'visual after-effect'. Prolonged viewing, termed 'adaptation', of a particular contour/texture shape causes other contour/texture shapes to appear distorted. By measuring the amount of distortion, a great deal can be learnt about how the brain processes contour/texture shapes. For example, if one takes a simple curved contour and adapts to it, a slightly more curved contour appears even more curved than otherwise. This effect is termed the 'curvature after-effect', and the fact that it happens at all tells us that there are neurons in the brain that are specialized for seeing curves. Thus the curvature after-effect can be used to uncover the details of how the brain sees curves. To take one example of how it can do this, my lab has shown that if the adaptation and test curves are a different colour the curvature after-effect is much reduced (i.e. there is less distortion) compared to if they are the same colour. This tells us that the neurons in the brain that detect curves are 'tuned' for colour, in other words some neurons respond to red curves, some green curves, some blue curves etc. Together with other experiments, my research therefore leads to a better understanding of how the brain sees the shapes of contours and textures.