With laptops and desktops, the dominant method of text entry is the full-sizekeyboard; now with the ubiquity of mobile devices like smartphones, two newwidely used methods have emerged: miniature touch screen keyboards andspeech-based dictation. It is currently unknown how these two modern methodscompare. We therefore evaluated the text entry performance of both methods inEnglish and in Mandarin Chinese on a mobile smartphone. In the speech inputcase, our speech recognition system gave an initial transcription, and thenrecognition errors could be corrected using either speech again or thesmartphone keyboard. We found that with speech recognition, the English inputrate was 3.0x faster, and the Mandarin Chinese input rate 2.8x faster, than astate-of-the-art miniature smartphone keyboard. Further, with speech, theEnglish error rate was 20.4% lower, and Mandarin error rate 63.4% lower, thanthe keyboard. Our experiment was carried out using Deep Speech 2, a deeplearning-based speech recognition system, and the built-in Qwerty or Pinyin(Mandarin) Apple iOS keyboards. These results show that a significant shiftfrom typing to speech might be imminent and impactful. Further research todevelop effective speech interfaces is warranted.
展开▼
