Modulation Enhancement of Temporal Envelopes for Increasing Speech Intelligibility in Noise

Koutsogiannaki, Maria; Stylianou, Yannis

doi:10.21437/interspeech.2016-500

Cited by 5 publications

(9 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thus, these results carry implications for our understanding of Lombard speech as occurring in natural communicative situations. They highlight the importance of speech enhancement techniques that are guided by naturally occurring speech, as for instance reported in Koutsogiannaki and Stylianou (2016) for clear speech.…”

Section: Discussionmentioning

confidence: 86%

“…Speech intelligibility greatly relies on these amplitude modulations, evident in the temporal envelope of speech (Drullman et al, 1994a;Shannon et al, 1995). In fact, enhancing the amplitude modulations in speech makes it more intelligible in noise (Koutsogiannaki & Stylianou, 2016), while filtering 6 amplitude modulations in the 1-9 Hz range out of the speech signal impairs intelligibility to a large degree (Drullman et al, 1994a(Drullman et al, , 1994bGhitza, 2012).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Enhanced amplitude modulations contribute to the Lombard intelligibility benefit: Evidence from the Nijmegen Corpus of Lombard Speech

Bosker

Cooke

2020

The Journal of the Acoustical Society of America

View full text Add to dashboard Cite

Speakers adjust their voice when talking in noise, which is known as Lombard speech. These acoustic adjustments facilitate speech comprehension in noise relative to plain speech (i.e., speech produced in quiet). However, exactly which characteristics of Lombard speech drive this intelligibility benefit in noise remains unclear. This study assessed the contribution of enhanced amplitude modulations to the Lombard speech intelligibility benefit by demonstrating that (1) native speakers of Dutch in the Nijmegen Corpus of Lombard Speech (NiCLS) produce more pronounced amplitude modulations in noise vs. in quiet; (2) more enhanced amplitude modulations correlate positively with intelligibility in a speech-in-noise perception experiment; (3) transplanting the amplitude modulations from Lombard speech onto plain speech leads to an intelligibility improvement, suggesting that enhanced amplitude modulations in Lombard speech contribute towards intelligibility in noise. Results are discussed in light of recent neurobiological models of speech perception with reference to neural oscillators phase-locking to the amplitude modulations in speech, guiding the processing of speech.

show abstract

Section: Discussionmentioning

confidence: 86%

Section: Introductionmentioning

confidence: 99%

Enhanced amplitude modulations contribute to the Lombard intelligibility benefit: Evidence from the Nijmegen Corpus of Lombard Speech

Bosker

Cooke

2020

The Journal of the Acoustical Society of America

View full text Add to dashboard Cite

show abstract

“…The effect was concentrated in higher frequency bands (around 1000 and 2000 Hz), which is in line with studies on artificial speech enhancement. For instance, Koutsogiannaki and Stylianou (2016) found that artificially decreasing the modulation depth in lower frequency regions (200-600 Hz) and increasing the modulation depth in higher frequencies (800-3000 Hz) enhanced speech-in-noise intelligibility.…”

Section: Discussionmentioning

confidence: 99%

“…These amplitude modulations, evident in the temporal envelope of speech, greatly contribute to speech intelligibility (Drullman et al, 1994;Ghitza, 2012;Shannon et al, 1995;Smith et al, 2002). Speech with more pronounced amplitude modulations is more intelligible in noise (Houtgast and Steeneken, 1985;Koutsogiannaki and Stylianou, 2016;Steeneken and Houtgast, 1980). Also, speakers who are intrinsically more intelligible than others show more pronounced low-frequency modulations in the temporal envelope (Bradlow et al, 1996).…”

Section: Introductionmentioning

confidence: 99%

Talkers produce more pronounced amplitude modulations when speaking in noise

Bosker

Cooke

2018

The Journal of the Acoustical Society of America

View full text Add to dashboard Cite

Speakers adjust their voice when talking in noise (known as Lombard speech), facilitating speech comprehension. Recent neurobiological models of speech perception emphasize the role of amplitude modulations in speech-in-noise comprehension, helping neural oscillators to “track” the attended speech. This study tested whether talkers produce more pronounced amplitude modulations in noise. Across four different corpora, modulation spectra showed greater power in amplitude modulations below 4 Hz in Lombard speech compared to matching plain speech. This suggests that noise-induced speech contains more pronounced amplitude modulations, potentially helping the listening brain to entrain to the attended talker, aiding comprehension.

show abstract

“…Many successful intelligibility enhancement methods modify the signal's features based on the noise masker ( [3], [4], [5]) while others exploit audio and signal properties ( [6]), [7]). Another family of algorithms that exploit human-like speech modifications has also been proposed ( [8], [9], [10], [11], [12]). These methods analyse casual speech and highly intelligible natural speech (Lombard speech [13], clear speech [14]) and modify the casual speech to reduce the feature differences between the two speaking styles.…”

Section: Introductionmentioning

confidence: 99%

Real-Time Modulation Enhancement of Temporal Envelopes for Increasing Speech Intelligibility

et al. 2017

Self Cite

View full text Add to dashboard Cite

In this paper, a novel approach is introduced for performing real-time speech modulation enhancement to increase speech intelligibility in noise. The proposed modulation enhancement technique operates independently in the frequency and time domains. In the frequency domain, a compression function is used to perform energy reallocation within a frame. This compression function contains novel scaling operations to ensure speech quality. In the time domain, a mathematical equation is introduced to reallocate energy from the louder to the quieter parts of the speech. This proposed mathematical equation ensures that the long-term energy of the speech is preserved independently of the amount of compression, hence gaining full control of the time-energy reallocation in real-time. Evaluations on intelligibility and quality show that the suggested approach increases the intelligibility of speech while maintaining the overall energy and quality of the speech signal.

show abstract

Modulation Enhancement of Temporal Envelopes for Increasing Speech Intelligibility in Noise

Cited by 5 publications

References 12 publications

Enhanced amplitude modulations contribute to the Lombard intelligibility benefit: Evidence from the Nijmegen Corpus of Lombard Speech

Enhanced amplitude modulations contribute to the Lombard intelligibility benefit: Evidence from the Nijmegen Corpus of Lombard Speech

Talkers produce more pronounced amplitude modulations when speaking in noise

Real-Time Modulation Enhancement of Temporal Envelopes for Increasing Speech Intelligibility

Contact Info

Product

Resources

About