Increasing Accessibility of Online Board Games to Visually Impaired People via Machine Learning and Textual/Audio Feedback: The Case of “Quantik”

Abstract

Playing board games is commonly recognized as an effective way to promote the integration and socialization of their participants. However, visually impaired people may encounter accessibility issues when playing online versions of board games, for instance, because such versions may have been designed having initially sighted people in mind. Given this premise, the aim of this work is to design an interface aimed to help visually impaired people play board games online, via an improved interaction with a normal or touch screen. This goal is achieved by means of automatic recognition of the portion of the screen one’s finger or the cursor is pointing to, its classification via machine learning, and the use of either textual or audio feedback. In this way, a visually impaired person could explore the screen in quite a natural way, obtaining information, e.g., about the positions of the various pieces on the board. As a case study, a preliminary version of the interface is developed to address accessibility of the online version of a carefully selected pure strategy abstract board game, namely “Quantik” from Gigamic.