Large Vocabulary Recognition of Wall Street Journal Sentences at Dragon Systems

Abstract

In this paper we present some of the algorithm improvements that have been made to Dragon's continuous speech recognition and training prograxns, improvements that have more than halved our error rate on the Resource Management task since the last SLS meeting in February 1991. We also report the "dry run" results that we have obtMned on the 5000-word speaker-dependent Wall Street Journal recognition task, and outline our overall research strategy and plans for the future.In our system, a set of output distributions, known as the set of PELs (phonetic elements), is associated with each phoneme. The HMM for a PIC (phoneme-in-context) is represented as a linear sequence of states, each having an output distribution chosen from the set of PELs for the given phoneme, and a (double exponential) duration distribution.In this paper we report on two methods of acoustic modeling and tr~ning. The first method involves generating a set of (unimodal) PELs for a given speaker by clustering the hypothetical frames found in the spectral models for that speaker, and then constructing speaker-dependent PEL sequences to represent each PIC. The "spectral model" for a PIC is simply the expected value of the sequence of frames that would be generated by the PIC. The second method represents the probability distribution for each parameter in a PEL as a mixture of a fixed set of unimodal components, the mixing weights being estimated using the EM algorithm. In both models we assume that the parameters axe statistically independent.We report results obtained using each of these two methods (RePELing/Respelling and univariate "tied mixtures") on the 5000-word closed-vocabulary verbalized punctuation version of the Wall Street Journal task.

Rapid match training for large vocabularies

“…Here we built new rapid match models for the 5K verbalized punctuation vocabulary for 5 of our 12 speakers, ranging from our worst performer to our best, and compared them to the original models which had already been adapted to each speaker. (For a description of our overall performance on the Wall Street Journal task, see the companion article [3].) The results are summarized in Table 1, which reports what percent of the time the correct word was included in the word candidate list returned by rapid match, as a function of the length of the list.…”

DARPA February 1992 pilot corpus CSR "dry run" benchmark test results

“…[5] and "It is encouraging that.., given there has not yet been any opportunity for parameter optimization." [6], and "The results, however, show promise and will require more rigorous testing." [7] and "This is a preliminary report demonstrating that.., was ported from a 1000-word task (ATIS) to a large vocabulary task (5000-word) task.., in three man weeks."…”

“…The four sites (CMU [5], Dragon Systems [6], MIT/LL [7] and SRI [8]) provided NIST with a total of 22 sets of results for a number of test sets and system configurations. The number of test set results provided by individual contractors ranged from 1 to 10.…”

Language identification via large vocabulary speaker independent continuous speech recognition

“…This recognition system has been described extensively elsewhere (see, for example, [5] and [6]). We review its basic properties here.…”

“…Our approach to language identification depends crucially on the existence of a large vocabulary continuous speech recognition system, so in order to test the feasibility of our language identification strategy, we turned to our primary LVCSR system, the Wall Street Journal recognizer developed under the ARPA SLS program. This recognition system has been described extensively elsewhere (see, for example, [5] and [6]). We review its basic properties here.…”