Spoken Input Research Articles

Microcomputer-based singing ability assessment and development

A new microcomputer-based system is described which has been developed for the assessment and development of singing ability. It makes use of a specially developed hardware interface which estimates the fundamental frequency of a sung or spoken input as the basis for (a) a measurement of vocal pitching accuracy (assessment), and (b) a vocal pitch display to provide visual feedback (development). Results are discussed which are based on a study carried out with the system in a British primary school, and they indicate that this system is effective in promoting singing development.

Speech recognition research at AT&T Bell Laboratories

At AT&T Bell Laboratories, a broad range of systems for speech recognition, depending on the intended application area, has been studied. These systems have been extensively studied for isolated word (and phrase) recognition for command and control applications where a single word (or phrase) suffices to effect some type of control over a system or service. Typical applications include call type recognition for automated dialing and call routing within an organization. Several systems have also been implemented for connected word recognition with vocabularies ranging from the digits up to several hundred words. Typical applications for this technology include order entry, credit card entry, and digit dialing. Most recently, several studies have been begun of large vocabulary (1000–50 000 words), continuous speech recognition. The intended application for this type of system is database management and access. Our primary focus has been in studying alternative representations of subword speech units. In particular, recognition systems based on phones (acoustic‐phonetic units), diphones, and acoustically defined units have been implemented and studied. Several methods of representing and accessing words from subword units including lexical access from a stored dictionary, explicit network representations of words, and derived acoustic lexicons have been studied. An investigation using a parser and a covering grammar to represent the constraints of the language and tack on the spoken input has also been shown. In this talk, the progress of each of these areas of speech recognition research is illustrated. Also discussed are the long‐term research goals and the steps intended to take to reach these goals.

Speaking backwards: A case study

A 31-year-old male philosophy professor (Prof. B) who can speak backwards fluently, at a normal rate, was studied. To define Prof. B's ability, recordings, spectrograms, and phonetic transcriptions were obtained. In an impromptu backwards monologue and in passages, sentences, and isolated words translated into backwards speech from written and spoken input sources in real time, Prof. B maintained the original word order but reversed the order of phonemes within each word. These reversals were not always phonetically complete, e.g., diphthongs were not reversed. Forward intonation was preserved. Although reversals were primarily phonologically based (e.g., silent consonants were rarely pronounced), there was a partial reliance on orthography even with spoken input (e.g., “xerox” was reversed as [ksɔriks], not [skariz]). Additional experiments investigated the processes contributing to Prof. B's special ability. In contrast to control subjects, Prof. B generally accomplished speech reversals “automatically.” However, reversal greatly increased Prof. B's recitation times when a processing load was introduced (e.g., in Z-to-A alphabetic recital). Finally, Prof. B possessed a much greater ability to identify words spoken backwards than did trained volunteers.

Active selection of items to be remembered: The role of timing

A series of experiments was carried out in which Ss were required to extract critical stimuli from a stream of nine spoken inputs, presented at various rates, and report on these after the presentation of each list. The critical items were normally digits at positions 2, 4, 6, and 8 in the input sequence. Subjects were required to employ either an “active” extraction strategy, aimed at achieving temporary storage only of items to be remembered, or a “passive” strategy, involving storage of all inputs with subsequent extraction of critical items. The initial experiment showed that the active strategy markedly improved performance efficiency as the presentation rate decreased; passive performance remained relatively stable. Experiments 2 and 3 indicated that the level of active performance was higher when critical items were categorically different from unwanted items. There were indications that this effect was independent of the effect of changes in the presentation rate. The final experiments in the series showed that when Ss were denied the opportunity of predicting the time of arrival of critical items active performance hardly benefitted from a reduction in rate. A “controlled activation” process is proposed as a basis for S's ability to modulate his state of alertness, so as to maximize receptivity for critical stimuli arriving at well-defined points in time.