Machine Learning, Robotic, and Toy Computing
Permanent URI for this collection
1 - 4 of 4
ItemNon-personal Data Collection for Toy User Interfaces( 2019-01-08)Toy-user-interfaces (ToyUI) are computing devices or peripherals that leverage interactivity and connectivity with other devices to promote physical and social play. ToyUI products may collect both personal and non-personal data (NPD) on their users. We propose nine data patterns for NPD collection as part of ToyUI design based on the study of 297 ToyUI items from both the literature and industry. In addition, we introduce a printed circuit board (PCB) used for rapid prototyping that enabled NPD data collection concerning both objects and users by gathering non-personal identification, positioning system, and motion tracking. We demonstrate the effectiveness of our hardware architecture by embedding it into two design scenarios, namely, closed rules and open-ended rules solutions. The objectives here are to assist the ToyUI makers in creating more meaningful play experiences while ensuring the privacy of children’s and their parents’ data.
ItemRoboTalk - Prototyping a Humanoid Robot as Speech-to-Sign Language Translator( 2019-01-08)Information science mostly focused on sign language recognition. The current study instead examines whether humanoid robots might be fruitful avatars for sign language translation. After a review of research into sign language technologies, a survey of 50 deaf participants regarding their preferences for potential reveals that humanoid robots represent a promising option. The authors also 3D-printed two arms of a humanoid robot, InMoov, with special joints for the index finger and thumb that would provide it with additional degrees of freedom to express sign language. They programmed the robotic arms with German sign language and integrated it with a voice recognition system. Thus this study provides insights into human–robot interactions in the context of sign language translation; it also contributes ideas for enhanced inclusion of deaf people into society.
ItemModeling Privacy Preservation in Smart Connected Toys by Petri-Nets( 2019-01-08)Children data privacy must be considered as integral and factored into the system design of Smart Connected Toy (SCT). The challenge is that SCTs are capable to gather significant amount volunteered and non-volunteered data, which lacks privacy considerations. It is imperative to adopt a modeling technique that autonomously preserves privacy and secure children’s data in SCT transactions. This paper surveys the current data flow modeling techniques, which most of them do not have elements to address the privacy of Personal Identifiable Information (PII). This paper shows a Petri-Net simulation which provides privacy assurance in order to minimize the risk of privacy violation of a child’s PII and related data.