Jazz guitar solos are improvised melody lines played on one instrument on top of a chordal accompaniment (comping). As the improvisation happens spontaneously, a reference score is non-existent, only a lead sheet. There are situations, however, when one would like to have the original melody lines in the form of notated music, see the Real Book. The motivation is either for the purpose of pract...

We present ``Transcriber'', a tool for assisting in the creation of speech corpora, and describe some aspects of its development and use. Transcriber was designed for the manual segmentation and transcription of long duration broadcast news recordings, including annotation of speech turns, topics and acoustic conditions. It is highly portable, relying on the scripting language Tcl/Tk with exten...

The goal of this study was to evaluate the reliability among transcribers of a standard prosodic labeling system under relatively optimal conditions of training, supervision, facilities, procedures, and extent of speaker familiarity. The ToBI (Tones and Break Indices) model for standard American English[7][1] was used in the study; break indices indicate the degree of junction between words, pi...

there is probably no need for extensive exploration to believe that movable type invention had a tremendous and historical impact on expansion of science and emergence of renaissance in europe, as tens of articles and research books have already been dedicated to this fact; but the real question which requires historical explanation is how the man before invention of printing had preserved his ...

This paper examines the standard string-matching approach to automated score following of an acoustic performance, looking at limitations imposed by both the score representation used and the matching approach. The approach is then extended through the incorporation of duration information and matchscores based on a model of transcriber error. The resulting score follower promises to allow comp...

This paper describes the Buckeye corpus of spontaneous American English speech, a 307,000-word corpus containing the speech of 40 talkers from central Ohio, USA. The method used to elicit and record the speech is described, followed by a description of the protocol that was developed to phonemically label what talkers said. The results of a test of labeling consistency are then presented. The c...

Phonetic transcriptions of spoken language corpora are not an exact written reproduction of the speech signal. They are influenced by a variety of factors such as the transcriber s native categorical perception. What remains unexplored is to what extent variation of perception within the same language exerts any influence on phonetic transcriptions. We report a case study of the labelling of vo...

In this paper, we propose a two-stage transcription task design for crowdsourcing with an automatic quality control mechanism embedded in each stage. For the first stage, a support vector machine (SVM) classifier is utilized to quickly filter poor quality transcripts based on acoustic cues and language patterns in the transcript. In the second stage, word level confidence scores are used to est...

In this paper, we offer an alternative to ToBI, the current de facto standard for machine-readable labelling of English prosody. We have three reasons for arguing that an alternative is needed. Firstly, the ToBI tone inventory is not maximally constrained. Inter-transcriber agreement for tone labels is relatively low and F0 contours generated from ToBI labels seem to be no closer to the origina...