Enhanced Captioning : Speaker Identification Using Graphical and Text-Based Identifiers

This thesis proposes a new technique for speaker identification in captioning using three identifiers: image, name and colour. This technique was implemented as a proofof-concept system called the Enhanced Captioning: Speaker Identification (EC: SID). This EC: SID was developed using participatory design and evaluated with people who are deaf or hard-of-hearing. This system evaluation used questionnaires and eye tracking methodologies, and the control was closed captioning, the existing system for North America. The results indicated that there is potential for using graphical and textbased identifiers for speaker identification. The placement of captioning or displaying the name of the speaker may not be effective for indicating who is speaking. The ability to customize these identifiers allows for changes in the content and different needs of users. Further design and evaluation is required to determine the long-term practicality of this graphical speaker identification technique.