Human navigation in complex spaces using landmarks and self-motion

复杂环境下的导航是人类认知系统的一项基本功能。然而，其背后的认知机制尚不清楚。我的研究计划的长期目标是充分了解人类如何发展和使用空间记忆，同时导航复杂的环境。 由于人类导航依赖于包括视觉和自我运动在内的多模态空间线索，我的研究计划的短期目标是研究人类如何使用视觉地标和自我运动来定位复杂空间中的自我和目标。我提出了三条研究路线。(1)这是一个共识的信念，远端地标主导方向线索，而近端地标主导位置线索，支持的实验中，人类的自我运动被打乱。我将提出并测试一个新的假设：当自我运动涉及导航，近端地标占主导地位的远端地标作为方向线索。(2)人们普遍认为，人们使用自我运动来不断更新他们的位置，并且使用自我运动的更新是自动的，独立于地标。然而，这一观点的经验证据并不一致。为了调和这种不一致，我将提出并测试一种新的理论，即人类使用自我运动来提供可能的位置，以便视觉搜索地标。(3)动物神经生理学表明，当老鼠爬上垂直的墙壁时，它们可以将它们的3d方向从地板延伸到墙壁。我将调查人类可以在多大程度上更新他们的3d标题。实证和理论研究的结果将实现我的短期目标，并为进一步的研究铺平道路，以实现我的长期目标。 方法论我将设计行为实验来测试三个研究领域中相互竞争的理论。在实验中，单个本科生（约3,000人，一半是男性）将在沉浸式虚拟环境中进行测试，我已经用它来研究人类导航20年了。他们将在走到测试位置之前学习物体的位置。他们将依靠自我运动或地标来定位记忆中的物体。它们在定位对象时的响应将用于计算它们对位置和航向的估计（例如，使用二维或三维回归）。通过研究不同线索对被试自我定位的影响，我将检验我的假设和替代假设。冲击该方案具有丰富的理论原创性和实验创新性，为人类导航领域和更广阔的领域做出了贡献。我最初的假设，人们更喜欢近端地标的方向与自我运动的导航远地标将完全改变现有的理论。我的创新范例，使用三维回归来计算人们在三维空间中的自我定位，将为研究人类的三维行为方向创造可能性。这些成果将对神经科学和导航人工智能产生更广泛的影响。

Navigation in complex environment is a fundamental function of the human cognitive system. However, the cognitive mechanisms underneath are not clear. The long-term goal of my research program is to fully understand how humans develop and use spatial memory while navigating complex environments. As human navigation relies on multimodal spatial cues including vision and self-motion, the short-term objective of my research program is to investigate how humans use visual landmarks and self-motion to locate self and goals in complex spaces. I propose three research lines. (1) It is a consensus belief that distal landmarks dominate directional cues while proximal landmarks dominate positional cues, supported by experiments in which humans' self-motion was disrupted. I will propose and test a new hypothesis: when self-motion is involved in navigation, proximal landmarks dominate over distal landmarks as directional cues. (2) It is generally agreed that people use self-motion to continuously update their locations and updating using self-motion is automatic, independent of landmarks. However, empirical evidence for this view is inconsistent. To reconcile the inconsistency, I will propose and test a new theory conjecturing that humans use self-motion to provide likely locations for visual search of the landmarks. (3) Animal neurophysiology demonstrates that rats can extend their 3d headings from the floor to the wall when they climb on vertical walls. I will investigate the extent to which humans can update their 3d headings. The empirical and theoretical outcomes will achieve my short-term objective and pave the way for further research to reach my long-term goal. Methodology. I will design behavioral experiments to test competing theories in the three research lines. In experiments, individual undergraduate students (~3,000, half male) will be tested in immersive virtual environments, which I have been used to study human navigation for 20 years. They will learn locations of objects before walking to the testing positions. They will locate remembered objects relying on self-motion or landmarks. Their response in locating objects will be used to calculate their estimates of position and heading (e.g., using a bi- or tri-dimension regression). By examining the roles of different cues on participants' self-location, I will test my hypotheses and alternative hypotheses. Impact. This proposal is rich in theoretical originality and experimental innovation, contributing to the field of human navigation and broader areas. My original hypothesis that people prefer proximal landmarks to distal landmarks for directions in navigation with self-motion will totally change the existing theory. My innovative paradigm of using tri-dimension regression to calculate peoples' self-location in 3d spaces will create the possibility to study humans' 3d headings behaviorally. The outcomes will have broader impacts on neuroscience and artificial intelligence of navigation.