Convolutive ICA-Based Forensic Speaker Identification Using Mel Frequency Cepstral Coefficients and Gaussian Mixture Models

Silveira, Matheus; Schroeder, Cezar; Costa, João Paulo; Oliveira, Celso José Bruno de; Apolinario, J.A.; Serrano, Antonio; Quintiliano, Paulo; Júnior, Rafael Sousa

doi:10.5769/j201301004

Cited by 7 publications

(10 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Using Blind Source Separation, Noise Suppression with the proposed VAD and Speaker Recognition in combination, the performance for poor audio conditions can be highly increased compared to state-of-the-art methods like [2]. The scheme we have proposed here has proven to work at low SNRs and with interference of various speakers.…”

Section: Discussionmentioning

confidence: 84%

“…According to [2], this method does not work at these SNR regimes. If no preprocessing is performed, the SR with noise with correlation above 0.7 is improved.…”

Section: B Resultsmentioning

confidence: 93%

“…The ICA approach and no pre-processing lead to the rather secure detection of both speakers when almost no noise is corrupting the records [2], whereas no preprocessing results in the good detection of one speaker above a certain SNR threshold. Below that threshold, none of the speakers is detected correctly.…”

Section: B Resultsmentioning

confidence: 99%

“…For this paper, we use Mel Frequency Cepstral Coefficients (MFCC), which in the past years have become one of the standard-approaches to this problem next to Linear Predictive Coding (LPC) and Local Discriminant Bases (LDB) [1]. Since MFCC is described in details in [2], [8], it shall not be treated in detail here. Afterwards, the histograms of the extracted features are matched with a distribution function (Box 3 of Figure 1).…”

Section: A Speaker Recognitionmentioning

confidence: 99%

“…al. have shown that using Blind Source Separation (BSS) techniques, the problem of identifying multiple speakers present in the same recording can be solved; however the proposed approach only works for high signal-to-noise ratio (SNR) regimes [2]. In this paper, we propose the use of a single-channel noise reduction (SCNR) scheme based on spectral subtraction [3] and a newly proposed Voice Activity Detector (VAD) in addition to separating the individual speakers using a BSS technique called convolutive Independent Component Analysis (ICA) [4], [5].…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Enhanced forensic multiple speaker recognition in the presence of coloured noise

Denk

Costa

Silveira

2014

2014 8th International Conference on Signal Processing and Communication Systems (ICSPCS)

View full text Add to dashboard Cite

Automatic speaker recognition has become a well-established technique for forensic applications. Since ambient recordings in such applications are obtained with hidden microphones far away from the sound sources, the performance of the speaker recognition can be severely degraded. In this paper, we propose an array signal processing method to compensate for these disturbances by spatially separating the present individual speakers and noise using convolutive Independent Component Analysis and applying a noise-suppression method based on spectral subtraction to the separated sound signals. A speaker recognition scheme based on Mel-Frequency Cepstral Coefficients and Gaussian Mixture Models is then applied to the separated and noise-cancelled signals. Our proposed pre-processing method dramatically increases the reliability of speaker recognition under such aggravated conditions and outperforms state-of-the-art solutions.

show abstract

Section: Discussionmentioning

confidence: 84%

“…According to [2], this method does not work at these SNR regimes. If no preprocessing is performed, the SR with noise with correlation above 0.7 is improved.…”

Section: B Resultsmentioning

confidence: 93%

Section: B Resultsmentioning

confidence: 99%

Section: A Speaker Recognitionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Enhanced forensic multiple speaker recognition in the presence of coloured noise

Denk

Costa

Silveira

2014

2014 8th International Conference on Signal Processing and Communication Systems (ICSPCS)

View full text Add to dashboard Cite

show abstract

End-to-end underwater acoustic source separation model based on EDBG-GALR

Yu,

Fan,

Cai

2024

Sci Rep

View full text Add to dashboard Cite

In practical work scenarios, it is common for multiple vessels to be present in close proximity within the same water area. In such cases, the acoustic signals emitted by these vessels often overlap, causing interference and reducing the accuracy of vessel identification. This paper proposes an improved GALR end-to-end underwater acoustic source separation algorithm, with ECA-DE (Efficient Channel Attention-Deep Encoder) and Bi-GRU (Bidirectional Gated Recurrent Unit), which consists of an encoder, separation blocks, and a decoder (EDBG-GALR). Addressing the issue of GALR encoder’s limited expressiveness due to its single-layer one-dimensional convolutional layer, we introduce a new deep encoder, ECA-DE, to enhance the encoder’s capability to process temporal signals and improve the efficiency of the separator’s input. Additionally, we integrate Bi-GRU into the GALR separation block’s local modeling module to enhance the local modeling ability of sequence features, thereby reducing the model’s computational and parameter requirements. Experimental results demonstrate that the proposed EDBG-GALR method can effectively separate mixed multi-target signals into multiple single-target signals, achieving a maximum scale-invariant signal-to-noise ratio (SI-SNR) improvement of 3.32 dB and a signal distortion ratio (SDR) improvement of 3.38 dB over baseline methods. These results highlight the practical applicability of EDBG-GALR in complex underwater environments.

show abstract

Low complexity blind separation technique to solve the permutation ambiguity of convolutive speech mixtures

Lima

Miranda

Costa

et al. 2016

2016 10th International Conference on Signal Processing and Communication Systems (ICSPCS)

View full text Add to dashboard Cite

Microphone arrays can be incorporated in several devices ranging from hearing aids and bioacoustic recording equipment to teleconference phones and forensic sound recorders in order to attenuate the interference of unwanted sounds. The separation of speech mixtures can be easily performed on the frequency domain independently for each frequency component. However, in order to combine the separated signals of each frequency component, the permutation ambiguity should be solved. The state-of-the-art technique relies on an iterative computation of the dispersions of the differences between the source profiles. In this paper, we propose a low complexity solution for the permutation ambiguity with similar accuracy based on one dispersion

show abstract

Convolutive ICA-Based Forensic Speaker Identification Using Mel Frequency Cepstral Coefficients and Gaussian Mixture Models

Cited by 7 publications

References 15 publications

Enhanced forensic multiple speaker recognition in the presence of coloured noise

Enhanced forensic multiple speaker recognition in the presence of coloured noise

End-to-end underwater acoustic source separation model based on EDBG-GALR

Low complexity blind separation technique to solve the permutation ambiguity of convolutive speech mixtures

Contact Info

Product

Resources

About