Perception of Material and Texture in Vision and Action

关于视觉系统如何处理、形成和塑造，有大量的研究。然而，关于物体特性，如表面纹理，指示重要的材料特性(例如，其粗糙度或柔软度)是如何被感知系统处理的，我们知道的要少得多。这是令人惊讶的，因为我们的感知系统很容易识别这些属性，但也令人惊讶，因为我们的许多行为决定，我们如何把握或与物体互动，取决于对材料属性的准确感知。人类成像研究表明，腹流中与处理形状、形状和颜色的区域不同的特定区域负责纹理辨别(Cant，Arnott&Goodale，2009)。然而，还不知道用什么视觉信息来估计材料的粗糙度或柔软度。最近，Giseel和Zaidi(2013)提出，视觉系统使用单目/基于图像的线索来估计图片中的粗糙度，这些线索与材料的3D结构有关。综上所述，这提出了几个有趣的问题：1)视觉系统如何为2D和3D刺激确定材料属性；2)材料属性的感知估计如何用于向视觉运动系统提供信息，反之亦然？3)视觉运动系统如何使用材料属性来调整其相应的动作？为了解决这些问题，我们将使用一系列技术，包括行为接触和掌握研究、知觉评级、材料属性的计算分析和神经心理学。在知觉实验中，已知与纹理的感知粗糙度有关的参数(即凸起的高度、密度、锐度和尖度)将在2D和3D刺激中独立地和系统地改变，并且将在心理物理实验中测量对知觉系统的影响。在视觉运动实验中，关于纹理的视觉和触觉信息将使用镜子设置来分离，这样参与者就可以抓住一个感觉不同(例如，更粗糙或更平滑)的物体(放在镜子后面)，而不是他们看到的物体(放在镜子前面)。对对象属性的感知估计将在对象交互之前和之后进行评估。这将使我们能够确定物质感知是否会随着触觉体验的变化而改变。这一点很重要，因为目前人们对视觉是否受到触觉体验的影响以及在多大程度上受到触觉体验的影响知之甚少。此外，计划进行一项神经心理学研究，以调查腹侧流损伤患者是否能够确定材料特性，如粗糙和柔软的纹理。如果这些估计如所建议的那样，依赖于从图片线索推断3D结构(与形状有关)，那么患者应该无法识别特定的材料属性，例如对于不熟悉的刺激，例如软的或粗糙的，即使在纹理辨别仍然可能的情况下。总而言之，这些实验将提供更多的了解，即知觉和视觉运动系统使用哪些线索来估计材料属性，以及这些信息是如何在人脑中处理的。

There is a vast amount of research on how the visual system processes form and shape. However, much less is known about how object properties, such as surface texture, that indicate important material properties (eg its roughness or softness), are processed by the perceptual system. This is surprising because our perceptual systems recognise these properties with ease, but also surprising because many of our behavioural decisions, how we grasp or interact with an object, depend on the accurate perception of material properties.Human imaging studies have shown that specific areas within the ventral stream, that are distinct from those processing shape, form and colour, are responsible for texture discrimination (Cant, Arnott, & Goodale, 2009). Yet, it is unknown which visual information is used to estimate the roughness or softness of a material. Recently, Giesel and Zaidi (2013) suggested that the visual system estimates roughness in pictures using monocular/image based cues which are related to the 3D structure of the material. Taken together, this raises several interesting questions: 1) Are there any differences in how the visual system determines material properties for 2D and 3D stimuli; 2) How is the perceptual estimation of material properties used to inform the visuomotor system and vice versa? and 3) How does the visuomotor system use material properties to adapt its actions accordingly?To address these questions, we will use a range of techniques, including behavioural reaching and grasping studies, perceptual ratings, computational analysis of material properties and neuropsychology. In the perceptual experiments, parameters that are known to relate to the perceived roughness of a texture (ie height, density, sharpness and pointedness of the bumps) will be varied independently and systematically in both 2D and 3D stimuli and the effects on the perceptual system will be measured in psychophysical experiments. In the visuomotor experiments, the visual and haptic information about texture will be dissociated using a mirror setup such that participants grasp an object (placed behind a mirror) that feels different (eg rougher or smoother) than the object they see (placed in front of a mirror). Perceptual estimates of object properties will be assessed prior to and after object interactions. This will allow us to determine whether material perceptions change depending on the haptic experience. This is important as currently little is known about if and to what extent vision is affected by tactile experience or vice versa. In addition, a neuropsychological study is planned to investigate if a patient with ventral stream damage that is known to be unable to perceive shape and form but has no problems discriminating between different textures and colours (Cavina-Pratesi, Kentridge, Heywood, & Milner, 2009), is able to determine material properties such as rough and soft textures. If these estimations, as suggested, rely on inferring 3D structure (relating to form) from pictorial cues, the patient should be unable to identify specific material properties such as soft or rough for unfamiliar stimuli even when texture discrimination is still possible. Together these experiments will provide an increased understanding of which cues the perceptual and visuomotor system use to estimate material properties and how this information is processed in the human brain.