Real-time transcription has been shown to be valuable in facilitating non-native speakers' comprehension in realtime communication. Automated speech recognition (ASR) technology is a critical ingredient for its practical deployment. This paper presents a series of studies investigating how the quality of transcripts generated by an ASR system impacts user comprehension and subjective evaluation. Experiments are first presented comparing performance across three different transcription conditions: no transcript, a perfect transcript, and a transcript with Word Error Rate (WER) =20%. We found 20% WER was the most likely critical point for transcripts to be just acceptable and useful. Then we further examined a lower WER of 10% (a lower bound for today's state-of-the-art systems) employing the same experimental design. The results indicated that at 10% WER comprehension performance was significantly improved compared to the no-transcript condition. Finally, implications for further system development and design are discussed.
Real-time transcription generated by automated speech recognition (ASR) technologies with a reasonably high accuracy has been demonstrated to be valuable in facilitating non-native speakers' comprehension in real-time communication. Besides errors, time delay often exists due to technical problems in automated transcription as well. This study focuses on how the time delay of transcription impacts non-native speakers' comprehension performance and user experience. The experiment design simulated a one-way computermediated communication scenario, where comprehension performance and user experiences in 3 transcription conditions (no transcript; perfect transcripts with a 2-second delay; and transcripts with a 10% word-error-rate and a 2-second delay) were compared. The results showed that the participants can benefit from the transcription with a 2-second time delay, as their comprehension performance in this condition was improved compared with the no-transcript condition. However, the transcription presented with delay was found to have negative effects on user experience. In the final part of the paper, implications for further system development and design are discussed.
Abstract. Measurements of user experience (UX) in traditional humancomputer interaction studies mostly rely on task performance and self-report data. Recent research has showed that physiological measures are good indicators of cognitive involvement and emotional arousal and are suggested being used as a complementary measure of UX. This paper reports a preliminary study to examine the possibility of including physiological measures in the UX evaluation process. In the experiment, participants' physiological responses, task performance and self-report data were collected and analyzed. It was found that physiological measures varied with task performance, as participants showed greater galvanic skin response (GSR) change in the failed tasks than that in the successful tasks. In addition, correlations were found between GSR and self-report data of user experience. The results demonstrated the potential value of physiological measures as a data source of user experience evaluation. However, further investigations involving variations in tasks and individual difference are required.
Brain-driven wheelchair control is an attractive application in theBrain-Computer Interface (BCI) field. In this research, wedesigned and validated a virtual wheelchair navigation systemcontrolled by our latest multiclass BCI Menu interface based on afast brain switch, which provides five commands: move forward,turn left, turn right, move backward, and stop. Preliminary resultshave shown that subjects can successfully control the wheelchairto hit all targets in the immersive virtual reality (VR)environment. This system proves an avenue to bridge the gapbetween simulation control in VR environments and real-lifewheelchair applications for mobility impairment.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.