Acoustical and Environmental Robustness in Automatic Speech by Alejandro Acero (auth.)

By Alejandro Acero (auth.)

The desire for automated speech attractiveness structures to be strong with recognize to adjustments of their acoustical setting has develop into extra largely liked lately, as extra platforms are discovering their manner into sensible functions. even if the difficulty of environmental robustness has got just a small fraction of the eye dedicated to speaker independence, even speech popularity structures which are designed to be speaker autonomous usually practice very poorly once they are proven utilizing a unique form of microphone or acoustical setting from the only with which they have been proficient. using microphones except a "close­ conversing" headset additionally has a tendency to significantly degrade speech acceptance -performance. Even in rather quiet workplace environments, speech is degraded through additive noise from fanatics, slamming doorways, and different conversations, in addition to by means of the results of unknown linear filtering coming up reverberation from floor reflections in a room, or spectral shaping via microphones or the vocal tracts of person audio system. Speech-recognition structures designed for long-distance cellphone traces, or functions deployed in additional antagonistic acoustical environments akin to motorized vehicles, manufacturing unit flooring, oroutdoors call for a ways greaterdegrees ofenvironmental robustness. There are a number of alternative ways of establishing acoustical robustness into speech popularity platforms. Arrays of microphones can be utilized to boost a directionally-sensitive approach that resists intelference from competing talkers and different noise assets which are spatially separated from the resource of the specified speech signal.

Show description

Read Online or Download Acoustical and Environmental Robustness in Automatic Speech Recognition, 1st Edition PDF

Similar acoustics & sound books

Multi-Platinum Pro Tools: Advanced Editing, Pocketing and Autotuning Techniques, Edition: Pap/Dvdr

Find out how a true expert makes use of seasoned instruments to make multi-platinum documents with this jam-packed, fast moving consultant. together with over three hundred colour illustrations, Multi-Platinum professional instruments takes you contained in the minds of 1 of the head seasoned instruments engineers within the enterprise, providing you with the talents you must prevail.

Photodetectors: An Introduction to Current Technology

This booklet has been written as a part of a brand new sequence of medical text-books being released by means of Plenum Publishing corporation restricted. The scope of the sequence is to check a selected subject in every one quantity, and likewise, to provide abstracts of an important references mentioned within the textual content. therefore permitting the reader to complement the knowledge contained inside this publication with out need to confer with many extra courses.

Applied Acoustics: Concepts, Absorbers, and Silencers for Acoustical Comfort and Noise Control: Alternative Solutions - Innovative Tools - Practical Examples

The writer supplies a finished review of fabrics and parts for noise keep an eye on and acoustical convenience. Sound absorbers needs to meet acoustical and architectural requisites, which fibrous or porous fabric by myself can meet. fundamentals and functions are established, with consultant examples for spatial acoustics, free-field try amenities and canal linings.

Audio Engineering 101: A Beginner's Guide to Music Production

Audio Engineering one zero one is a true global advisor for beginning out within the recording undefined. in case you have the dream, the tips, the tune and the creativity yet have no idea the place to begin, then this booklet is for you! full of functional suggestion on easy methods to navigate the recording global, from an writer with first-hand, real-life event, Audio Engineering one zero one might help you reach the intriguing, yet tricky and complicated, track undefined.

Extra info for Acoustical and Environmental Robustness in Automatic Speech Recognition, 1st Edition

Example text

The distortion measure estimates the degree of proximity of two vectors. An input vector is mapped to a symbol of this alphabet by choosing the closest codebook vector. In SPHINX the distortion measure used is the Euclidean distance. The version of SPHINX that we used for this work has three different codebooks: one for the cepstrum, one for the first difference of cepstral 14g is the zeroth order cepstral coefficient. 20 ACOUSTICAL AND ENVIRONMENTAL ROBUSTNESS vectors and the last one for power and the first difference of the power.

Finally, additional databases with different sets of microphones were recorded in stereo. The average spectra, SNR calculations and baseline performance were computed for all of them. Frequency Domain Processing In this chapter we review some of the techniques proposed in the literature to deal with the problem of robustness to noise and tilt in the spectrum. These techniques were adapted to our present system and modified when necessary. Several approaches have been tried in conjunction with our census database and the SPHINX system.

He suggested that this technique could also be used to increase the robustness of the system in noisy environments. 6. Other Techniques There are a number of different approaches to the problem of speech recognition systems that are robust to noise. Some of these other techniques include neural networks and the use of microphone arrays. Tamura and Waibel [80] suggested the use of a neural network for speech enhancement that was trained to minimize the difference between noisy and clean waveforms.

Download PDF sample

Rated 4.40 of 5 – based on 22 votes