Detection and Processing of Information in Three-Dimensional Space

我的研究计划的长期目标是理解人类空间能力背后的基本感知原则。利用虚拟现实(VR)技术和各种真实世界测试范式，我们将进一步进行以下三个方面的研究：1)运动中的多感觉整合；2)对即将发生的碰撞的视觉控制；3)在近个人和非个人空间中感知视觉事件。*我们的一线研究重点是人类大脑在移动过程中如何解释和整合视觉(光流)和基于身体(本体感觉、前庭)的信息。为了单独评估每个感觉线索的作用，并更好地了解这些线索是如何相互作用和相互整合形成单一知觉的，我们使用了几种涉及自我运动(如骑自行车和走路)的新颖实验范式来创造线索冲突情景。这个第一个研究项目的下一个目标将集中在进一步研究和量化自我运动过程中视觉和基于身体的信息之间发生的动态交互作用。通过使用贝叶斯模型对经验结果进行建模，我们将能够更好地量化以前的感觉体验如何影响当前的感觉处理。这一系列研究不仅将帮助我们理解感知行驶距离的机制，还将提供对定义大脑如何组合多个信息源的一般原理的洞察。*虽然我们的第一线研究侧重于对观察者在环境中过去移动的程度的多感官处理，而我们的第二线研究专注于观察者在任一物体运动(例如，迎面而来的汽车)期间拦截环境中的物体之前的未来碰撞的视觉处理，以及由于观察者的自我运动(例如，行走和驾驶)而经历的视觉环境的运动。我们最近的研究提供了明确的经验证据，表明观察者使用光学变量tau来判断即将发生的碰撞。在我们的下一个赠款期间，我们将进行心理物理研究，探索更复杂的运动形式的处理机制。我们将比较人类在物体运动、自我运动或两者都有的情况下的表现。除了对运动的感知，我们还将在我们的高端驾驶模拟器(配备眼球跟踪)上，在模拟驾驶任务中检验复杂、自然的运动处理和碰撞避免形式的视觉运动控制。*我们的第三线研究将跟进我们最近的令人兴奋的发现，即对呈现在近(个人周围)和远(个人以外)观看距离的视觉目标进行差异处理，即使在近距离和远距离观看条件下的视觉刺激在视角和亮度方面是匹配的。我们将继续研究这种观看距离效应的潜在机制。到目前为止，我们的试点数据表明，近视觉空间和远视觉空间可能分别由背侧和腹侧视觉流进行不同处理。*我们虚拟现实实验室的设置为三维、多感官、动态、生态有效的刺激呈现提供了很好的实验控制。这个研究项目不仅让我探索了视觉运动控制的基本感知原理，而且还可以用于设计更好的标准化测试和驾驶员/飞行员的培训/康复，以及汽车、路标和驾驶/飞行模拟器的设计。

The long-term aim of my research programme is to understand the fundamental perceptual principles underlying human spatial abilities. Using virtual reality (VR) technology and various real world testing paradigms, we will further pursue the following three lines of research 1) multisensory integration in locomotion and 2) visual control of impending collision; 3) perception of visual events in near peripersonal and extrapersonal space.* Our first line of research focuses on how the human brain interprets and integrates visual (optic flow) and body-based (proprioceptive, vestibular) information during locomotion. In order to assess the role of each sensory cue individually and gain a better understanding of how these cues interact with and are integrated with one another to form a single perception, we have used several novel experimental paradigms involving self-motion (such as bicycling and walking) to create cue conflict scenarios. The next goal of this first research program will concentrate on further investigating and quantifying the dynamic interactions that occur between visual and body-based information during self-motion. By modeling the empirical findings using Bayesian model we will be better equipped to quantify how previous sensory experiences affect current sensory processing. This line of research will not only help us understand the mechanisms involved in perception of travelled distance, but also provides insight into the general principles that define how the brain combines multiple sources of information.* While our first line of research focus on the multisensory processing of the extent of PAST movement of the observers in the environment, our second line of research concentrate on the visual processing of FUTURE collision before the observers intercept objects in the environment during either object motion (e.g., an oncoming car), and the motion of the visual environment experienced as a result of the observer's self-motion (e.g., walking and driving). Our recent studies provided clear empirical evidence demonstrating that observers use optical variable tau to make judgement of impending collision. In our next grant period we will conduct psychophysical studies exploring the mechanisms of processing for more complicated forms of motion. We will compare human performance under object motion, self-motion, or both. In addition to perception of motion, we will also examine visual motor control in complex, naturalistic forms of motion processing and collision avoidance in simulated driving tasks on our high-end driving simulator (equipped with eye tracking). * Our third line of research will follow up our recent exciting discoveries of differential processing of visual targets presented at near (peripersonal) and far (extrapersonal) viewing distances, even though visual stimuli in near and far viewing conditions were matched for visual angle and luminance. We will continue to examine the underlying mechanisms for such a viewing distance effect. Our pilot data so far suggest that near and far visual spaces might be differentially processed by dorsal and ventral visual streams respectively.* The setups in our VR laboratory afford great experimental control over three-dimensional, multisensory, dynamic, ecologically valid, stimulus presentations. This research program has not only allowed me to explore fundamental perceptual principles underlying visuomotor control, but it can also be employed in the design of better standardized tests and training/rehabilitation for drivers/pilots and the design of car, road sign and driving/flight simulators.