The aim of the project reported in this paper is to investigate the potential for 'SpeakerID' to improve the effectiveness of communication aids. SpeakerID is software that detects who you are talking to. This would allow a communication aid to 'know' who was talking to the person using the communication aid and adapt its interface accordingly.
SpeakerID is an example of the potential role of 'context' within aided communication; another example is location, i.e. knowing where you are. Use of contextual information gives the potential to improve the simplicity and efficacy of these systems (although, implemented badly, it also has potential to do the opposite).
This project, which was underway at the time of writing, aims to produce a proof of concept implementation of SpeakerID which could be connected to VOCA software. The proof of concept device will be used as an 'artefact' to stimulate discussion and better understand the potnetial user requirements for this kind of technique.
At the time of writing, the project was still underway and the prototype was being developed. This session will be used to stimulate discussion around the potential for 'contextual information' such as SpeakerID.
The use of contextual information is relatively often mentioned in the AAC literature, but little or no work appears to have been to find out if this actually useful or needed, what the 'most important' piece of context is in a conversation, or how it can be integrated into VOCA software which appears to not be set up to 'accept' this information. This project aims to prototype an 'artefact' to investigate these areas.