An FFT-Based Companding Front End for Noise-Robust Automatic Speech Recognition
We describe an FFT-based companding algorithm for preprocessing speech before recognition. The algorithm mimics tone-to-tone suppression and masking in the auditory system to improve automatic speech recognition performance in noise. Moreover, it is also very computationally efficient and suited to...
| Main Authors: | , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
Hindawi Publishing Corporation
2011
|
| Online Access: | http://hdl.handle.net/1721.1/67033 https://orcid.org/0000-0003-0384-3786 |
| _version_ | 1826203666846056448 |
|---|---|
| author | Raj, Bhiksha Turicchia, Lorenzo Schmidt-Nielsen, Bent Sarpeshkar, Rahul |
| author2 | Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science |
| author_facet | Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Raj, Bhiksha Turicchia, Lorenzo Schmidt-Nielsen, Bent Sarpeshkar, Rahul |
| author_sort | Raj, Bhiksha |
| collection | MIT |
| description | We describe an FFT-based companding algorithm for preprocessing speech before recognition. The algorithm mimics tone-to-tone suppression and masking in the auditory system to improve automatic speech recognition performance in noise. Moreover, it is also very computationally efficient and suited to digital implementations due to its use of the FFT. In an automotive digits recognition task with the CU-Move database recorded in real environmental noise, the algorithm improves the relative word error by 12.5% at -5 dB signal-to-noise ratio (SNR) and by 6.2% across all SNRs (-5 dB SNR to +5 dB SNR). In the Aurora-2 database recorded with artificially added noise in several environments, the algorithm improves the relative word error rate in almost all situations. |
| first_indexed | 2024-09-23T12:41:10Z |
| format | Article |
| id | mit-1721.1/67033 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T12:41:10Z |
| publishDate | 2011 |
| publisher | Hindawi Publishing Corporation |
| record_format | dspace |
| spelling | mit-1721.1/670332022-09-28T09:24:39Z An FFT-Based Companding Front End for Noise-Robust Automatic Speech Recognition Raj, Bhiksha Turicchia, Lorenzo Schmidt-Nielsen, Bent Sarpeshkar, Rahul Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Massachusetts Institute of Technology. Research Laboratory of Electronics Turicchia, Lorenzo Turicchia, Lorenzo Sarpeshkar, Rahul We describe an FFT-based companding algorithm for preprocessing speech before recognition. The algorithm mimics tone-to-tone suppression and masking in the auditory system to improve automatic speech recognition performance in noise. Moreover, it is also very computationally efficient and suited to digital implementations due to its use of the FFT. In an automotive digits recognition task with the CU-Move database recorded in real environmental noise, the algorithm improves the relative word error by 12.5% at -5 dB signal-to-noise ratio (SNR) and by 6.2% across all SNRs (-5 dB SNR to +5 dB SNR). In the Aurora-2 database recorded with artificially added noise in several environments, the algorithm improves the relative word error rate in almost all situations. 2011-11-16T13:45:46Z 2011-11-16T13:45:46Z 2007-06 2006-11 2011-09-23T17:09:42Z Article http://purl.org/eprint/type/JournalArticle 1687-4714 1687-4722 http://hdl.handle.net/1721.1/67033 EURASIP Journal on Audio, Speech, and Music Processing. 2007 Jun 26;2007(1):065420 https://orcid.org/0000-0003-0384-3786 en http://dx.doi.org/10.1155/2007/65420 EURASIP Journal on Audio, Speech, and Music Processing Creative Commons Attribution http://creativecommons.org/licenses/by/2.0 et al.; licensee BioMed Central Ltd. application/pdf Hindawi Publishing Corporation |
| spellingShingle | Raj, Bhiksha Turicchia, Lorenzo Schmidt-Nielsen, Bent Sarpeshkar, Rahul An FFT-Based Companding Front End for Noise-Robust Automatic Speech Recognition |
| title | An FFT-Based Companding Front End for Noise-Robust Automatic Speech Recognition |
| title_full | An FFT-Based Companding Front End for Noise-Robust Automatic Speech Recognition |
| title_fullStr | An FFT-Based Companding Front End for Noise-Robust Automatic Speech Recognition |
| title_full_unstemmed | An FFT-Based Companding Front End for Noise-Robust Automatic Speech Recognition |
| title_short | An FFT-Based Companding Front End for Noise-Robust Automatic Speech Recognition |
| title_sort | fft based companding front end for noise robust automatic speech recognition |
| url | http://hdl.handle.net/1721.1/67033 https://orcid.org/0000-0003-0384-3786 |
| work_keys_str_mv | AT rajbhiksha anfftbasedcompandingfrontendfornoiserobustautomaticspeechrecognition AT turicchialorenzo anfftbasedcompandingfrontendfornoiserobustautomaticspeechrecognition AT schmidtnielsenbent anfftbasedcompandingfrontendfornoiserobustautomaticspeechrecognition AT sarpeshkarrahul anfftbasedcompandingfrontendfornoiserobustautomaticspeechrecognition AT rajbhiksha fftbasedcompandingfrontendfornoiserobustautomaticspeechrecognition AT turicchialorenzo fftbasedcompandingfrontendfornoiserobustautomaticspeechrecognition AT schmidtnielsenbent fftbasedcompandingfrontendfornoiserobustautomaticspeechrecognition AT sarpeshkarrahul fftbasedcompandingfrontendfornoiserobustautomaticspeechrecognition |