Date of Award

5-1-2019

Document Type

Thesis (Undergraduate)

Department or Program

Department of Computer Science

First Advisor

Xia Zhou

Abstract

Uncanny valley is a theorized psychological phenomenon, which captures a non-monotonic relationship between an entity's anthropomorphic level and the shinwakan (affinity) its viewers feel toward the entity. According to the theory, viewers feel a stronger affinity to an anthropomorphic entity as its level of human likeness increases until it reaches a certain point where that affinity is brought to a sudden drop. This valley, although frequently observed, still remains not well understood or explained. That said, most studies purport to present an explanation to the valley in context of robotics or computer-generated images portrayed on 2D surfaces, but it is unclear whether these explanations are applicable to different platforms that offer a completely different user experience. Hence, this study attempts to explore the uncanny valley in context of the augmented reality. AR overlays texts and images on top of our visual feed to interactively blend the real and the virtual world. With its immersive and interactive nature, AR may potentially offer us a new perspective in understanding uncanny valley. The rules that we believed governed the uncanny valley in robotics and previous 2D platforms may no longer be applicable to AR. Consequently, this study observes a relational trend between entities uncanniness and human likeness using Microsoft Hololens via the user survey and biometrics data to offer viable explanations to how we should interpret the uncanny valley in AR setting.

Comments

Originally posted in the Dartmouth College Computer Science Technical Report Series, number TR2019-866.

Download
COinS