Speaker identification from shouted speech: Analysis and compensation

Hanilçi, Cemal; Kinnunen, Tomi; Saeidi, Rahim; Pohjalainen, Jouni; Alku, Paavo; Ertaş, Figen

doi:10.1109/icassp.2013.6639228

Cited by 13 publications

(12 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A number of stress compensation strategies were also formulated, which resulted in significant improvement in ASR (Hansen and Clements, 1989;Hansen, 1996). Later studies considered individual speech modes such as whisper (Fan and Hansen, 2011) and shout (Hanilci et al, 2013) in the context of speaker identification and proposed different compensation strategies to normalize mismatch.…”

Section: Introductionmentioning

confidence: 99%

Analysis of human scream and its impact on text-independent speaker verification

Hansen

Nandwana

Shokouhi

2017

The Journal of the Acoustical Society of America

View full text Add to dashboard Cite

Scream is defined as sustained, high-energy vocalizations that lack phonological structure. Lack of phonological structure is how scream is identified from other forms of loud vocalization, such as “yell.” This study investigates the acoustic aspects of screams and addresses those that are known to prevent standard speaker identification systems from recognizing the identity of screaming speakers. It is well established that speaker variability due to changes in vocal effort and Lombard effect contribute to degraded performance in automatic speech systems (i.e., speech recognition, speaker identification, diarization, etc.). However, previous research in the general area of speaker variability has concentrated on human speech production, whereas less is known about non-speech vocalizations. The UT-NonSpeech corpus is developed here to investigate speaker verification from scream samples. This study considers a detailed analysis in terms of fundamental frequency, spectral peak shift, frame energy distribution, and spectral tilt. It is shown that traditional speaker recognition based on the Gaussian mixture models-universal background model framework is unreliable when evaluated with screams.

show abstract

Section: Introductionmentioning

confidence: 99%

Analysis of human scream and its impact on text-independent speaker verification

Hansen

Nandwana

Shokouhi

2017

The Journal of the Acoustical Society of America

View full text Add to dashboard Cite

show abstract

“…Speaker recognition process encompasses three terms identification, verification and diarization. In automatic speaker identification (Figure 1), there is no priori identity claim, the system decides who the person is, or the person is known or unknown [2] [12]. While automatic speaker verification (Figure 2) involves the use of a machine to verify a person's claimed identity from his voice [6,7,12].…”

Section: Speaker Recognitionmentioning

confidence: 99%

Smart Voice Search Engine

Sarhan¹

2014

IJCA

View full text Add to dashboard Cite

For years and years the search engine researchers focus their efforts on having more accurate and faster search engines. This was more than enough in the past but with the appearance of smart phones the idea of having everything smart became dominant. In this paper we tried to share the dream of having a domain independent search engine and not only an ordinary one but a smart search by voice engine which searches user speech automatically without the user 's request and provide him with evidence on his speech, this engine was called SVSE. Through the paper we will introduce the proposed system in details explaining each part of it and finally discussing the difficulties it faces. General Termsnatural Language Processing, speech Recognition.

show abstract

“…Another strategy, which has not been widely explored, is to use feature mapping. A recent study showed that such an approach can be helpful in speaker identification scenarios when the input presented is shouted speech [10]. For feature mapping, neural networks and Gaussian mixture models have been widely used in the voice conversion and voiced speech reconstruction (from whispered speech) literature [11,12,13].…”

Section: Introductionmentioning

confidence: 99%

Feature mapping, score-, and feature-level fusion for improved normal and whispered speech speaker verification

Sarria-Paja

Senoussaoui

O’Shaughnessy

et al. 2016

2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

In this paper, automatic speaker verification using normal and whispered speech is explored. Typically, for speaker verification systems with varying vocal effort inputs, standard solutions such as feature mapping or addition of data during parameter estimation (training) and enrollment stages result in a trade-off between accuracy gains with whispered test data and accuracy losses (up to 70% in equal error rate, EER) with normal test data. To overcome this shortcoming, this paper proposes two innovations. First, we show the complementarity of features derived from AM-FM models over conventional mel-frequency cepstral coefficients, thus signalling the importance of instantaneous phase information for whispered speech speaker verification. Next, two fusion schemes are explored: score-and feature-level fusion. Overall, we show that gains as high as 30% and 84% in EER can be achieved for normal and whispered speech, respectively, using featurelevel fusion.

show abstract

Speaker identification from shouted speech: Analysis and compensation

Cited by 13 publications

References 10 publications

Analysis of human scream and its impact on text-independent speaker verification

Analysis of human scream and its impact on text-independent speaker verification

Smart Voice Search Engine

Feature mapping, score-, and feature-level fusion for improved normal and whispered speech speaker verification

Contact Info

Product

Resources

About