Juicer: A Weighted Finite-State Transducer Speech Decoder

Abstract

Abstract. A major component in the development of any speech recognition system is the decoder. As task complexities and, consequently, system complexities have continued to increase the decoding problem has become an increasingly significant component in the overall speech recognition system development effort, with efficient decoder design contributing to significantly improve the trade-off between decoding time and search errors. In this paper we present the "Juicer" (from transducer ) large vocabulary continuous speech recognition (LVCSR) decoder based on weighted finite-State transducer (WFST). We begin with a discussion of the need for open source, state-of-the-art decoding software in LVCSR research and how this lead to the development of Juicer, followed by a brief overview of decoding techniques and major issues in decoder design. We present Juicer and its major features, emphasising its potential not only as a critical component in the development of LVCSR systems, but also as an important research tool in itself, being based around the flexible WFST paradigm. We also provide results of benchmarking tests that have been carried out to date, demonstrating that in many respects Juicer, while still in its early development, is already achieving state-of-the-art. These benchmarking tests serve to not only demonstrate the utility of Juicer in its present state, but are also being used to guide future development, hence, we conclude with a brief discussion of some of the extensions that are currently under way or being considered for Juicer.

Enhanced Phone Posteriors for Improving Speech Recognition Systems

“…The specifications of databases, regular MLP posteriors, and enhanced posterior estimation are the same as mentioned in Section 4.2. We have used JUICER [35] as the hybrid decoder. In case of Numbers database, phones are modeled with 5 states in the decoder.…”

A Generalized Dynamic Composition Algorithm of Weighted Finite State Transducers for Large Vocabulary Speech Recognition

“…A trigram LM, with 19979 unigrams, 3484372 bigrams and 2949590 trigrams, was used to test the 20k development test set "si dt 20" from WSJ1 database, consisting of 503 utterances. The experiment was carried out using Juicer [12,13], which is a WFST-based LVCSR decoder developed here at IDIAP. Figure 3 shows the word error rate (WER) against the real-time factor (RTF) of different approaches.…”

Accelerate Acoustic Likelihood Computations on GPU for Speech Recognition

“…In the field of decoder [8] speed for LVCSR, a major achievement is the architecture of weighted finite-state transducers (WFST) [9][10][11]. WFST have an improvement on the speed than the tree-decoder.…”