Speech recognition using HMMs with quantized parameters
نویسنده
چکیده
In this paper we describe the structure and examine the performance of a recognition engine based on hidden Markov models (HMMs) with quantized parameters (qHMM). The main goal of qHMMs is to enable a low complexity implementation without sacrificing the classification performance. In the tests with a whole word digit dialler engine and a phoneme based isolated word recognizer we managed to preserve the performance of unquantized HMMs with qHMMs having as little as 5 bit for a mean component and 3 bit for a variance component.
منابع مشابه
Speaker adaptation of quantized parameter HMMs
This paper extends the evaluation of Hidden Markov Models with quantized parameters (qHMM) presented in [5] to the case of speaker adaptive training. In speaker-independent speech recognition tasks, qHMMs were found to provide a similar performance as the original continuous density HMMs (CDHMM) with substantially reduced memory requirements. In this paper, we propose a Bayesian type of adaptat...
متن کاملLow memory acoustic models for HMM based speech recognition
In this paper, we propose a new approach to reduce the memory footprint of HMM based ASR systems. The proposed method involves three steps. Starting from the continuous density HMMs, mixture variances are tied using k-means based vector quantization. Next, the reestimation of the resulted models is performed with tied variances. Finally, scalar quantization is performed for the mean components ...
متن کاملQuantized HMMs for low footprint text-to-speech synthesis
This paper proposes the use of Quantized Hidden Markov Models (QHMMs) for reducing the footprint of conventional parametric HMM-based TTS system. Previously, this technique was successfully applied to automatic speech recognition in embedded devices without loss of recognition performance. In this paper we investigate the construction of different quantized HMM configurations that serve as inpu...
متن کاملComparison of low footprint acoustic modeling techniques for embedded ASR systems
In this paper we compare the performance of speech recognition systems based on hidden Markov models (HMM) with quantized parameters (qHMMs) and subspace distribution clustering hidden Markov models (SDCHMMs). Both of these HMM types provide similar performance as continuous density HMMs, but with significantly reduced memory requirements (approximately 90% less memory was needed to store the H...
متن کاملPLP coefficients can be quantized at 400 bps
Previous work in wireless speech recognition has focused on two methods, namely, quantizing recognition features (e.g. MFCC) or performing recognition using speech coding parameters (e.g. LPC). All of this previous research assumes that the communication channel is only large enough to transmit either speech coding parameters or speech recognition parameters. By contrast, we propose that the sp...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2000