Speaker Adaptation Using Evo Transfor
نویسنده
چکیده
This paper presents a technique to adapt HMMs to new speakers by using Genetic Algorithms (GAs) in unsupervised mode. The implementation requirements of GAs, such as genetic operators and objective function, have been chosen in order to give more reliability to a global linear transformation matrix. By implementing a ‘survival of the fittest’ strategy, the proposed GA-MLLR approach allows to maintain and manipulate a population of a wide range of solutions.Experiments have been performed on ARPA-RM and TIMIT databases using a triphones HMM-based system. Results show that from a new speaker, significant decrease of word error rate which can reach 6% for a particular speaker, has been achieved by the evolutionary approach, compared to the conventional MLLR-based adaptation method.
منابع مشابه
Speaker Adaptation in Continuous Speech Recognition Using MLLR-Based MAP Estimation
A variety of methods are used for speaker adaptation in speech recognition. In some techniques, such as MAP estimation, only the models with available training data are updated. Hence, large amounts of training data are required in order to have significant recognition improvements. In some others, such as MLLR, where several general transformations are applied to model clusters, the results ar...
متن کاملSpeaker Adaptation in Continuous Speech Recognition Using MLLR-Based MAP Estimation
A variety of methods are used for speaker adaptation in speech recognition. In some techniques, such as MAP estimation, only the models with available training data are updated. Hence, large amounts of training data are required in order to have significant recognition improvements. In some others, such as MLLR, where several general transformations are applied to model clusters, the results ar...
متن کاملLong term on-line speaker adaptation for large vocabulary dictation
On-line speaker adaptation is desirable for speech recognition dictation applications, because it o ers the possibility to improve the system with the speaker-speci c data obtained from the user. Since the user will work with such a device over a long period, for a dictation system the long term adaptation performance is more important than the adaptation speed. In contrast to speaker-dependent...
متن کاملAn online incremental speaker adaptation method using speaker-clustered initial models
We previously proposed an incremental speaker adaptation method combined with automatic speaker-change detection for broadcast news transcription where speakers change frequently and each of them utters a series of several sentences. In this method, the speaker change is detected using speaker-independent and speaker-adaptive Gaussian mixture models (GMMs). Both phone HMMs and GMMs are incremen...
متن کاملUsing genetic algorithms for rapid speaker adaptation
This paper proposes two new approaches to rapid speaker adaptation of acoustic models by using genetic algorithms. Whereas conventional speaker adaptation techniques yield adapted models which represent local optimum solutions, genetic algorithms are capable to provide multiple optimal solutions, thereby delivering potentially more robust adapted models. We have investigated two different strat...
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 2006