A Robust Text Dependent Speaker Identification Using Neural Responses from the Model of the Auditory System

Khalil, Md. Ibrahim; Mamun, Nursadul; Akter, Khadija

doi:10.1109/ecace.2019.8679215

Cited by 4 publications

(3 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The goal of speech separation is to separate a set of speech signals from a set of mixed signals. In this paper, we focus on single-microphone, speaker-independent source separation, which has many potential applications such as: digital hearing aids [5,6], automatic speech recognition [7,8,9], speaker diarization [10,11], emotion recognition [12,13], speaker verification and identification [14,15]. Researchers have proposed many methods to solve the speech separation task; these methods can be divided into two categories: (1) unsupervised, and (2) supervised.…”

Section: Introductionmentioning

confidence: 99%

Probabilistic Permutation Invariant Training for Speech Separation

2019

View full text Add to dashboard Cite

Single-microphone, speaker-independent speech separation is normally performed through two steps: (i) separating the specific speech sources, and (ii) determining the best output-label assignment to find the separation error. The second step is the main obstacle in training neural networks for speech separation. Recently proposed Permutation Invariant Training (PIT) addresses this problem by determining the output-label assignment which minimizes the separation error. In this study, we show that a major drawback of this technique is the overconfident choice of the output-label assignment, especially in the initial steps of training when the network generates unreliable outputs. To solve this problem, we propose Probabilistic PIT (Prob-PIT) which considers the output-label permutation as a discrete latent random variable with a uniform prior distribution. Prob-PIT defines a log-likelihood function based on the prior distributions and the separation errors of all permutations; it trains the speech separation networks by maximizing the loglikelihood function. Prob-PIT can be easily implemented by replacing the minimum function of PIT with a soft-minimum function. We evaluate our approach for speech separation on both TIMIT and CHiME datasets. The results show that the proposed method significantly outperforms PIT in terms of Signal to Distortion Ratio and Signal to Interference Ratio.

show abstract

Section: Introductionmentioning

confidence: 99%

Probabilistic Permutation Invariant Training for Speech Separation

2019

View full text Add to dashboard Cite

show abstract

“…surroundings and receive a reward for good performance and penalties for bad behaviors, which over time enables maximization of their cumulative rewards and solving the problem as much as possible [22].…”

Section: Artificial Neural Network (Ann)mentioning

confidence: 99%

Automatic Speaker Identification System based MLP Network and Deep Learning in the Presence of Severe Interference

Shafik¹,

Sedik

Khalaf

et al. 2023

Preprint

View full text Add to dashboard Cite

This paper proposes two models for Automatic Speaker Identification (ASI) task in the presence of interference. The first one uses the Normalized Pitch Frequency (NPF) and Mel- Frequency Cepstral Coefficients (MFCCs) as extracted features and multi-layer perceptron (MLP) as a classifier. In this model, we investigate the utilization of the Discrete Transform (DT) such as Discrete Wavelet Transform (DWT), Discrete Cosine Transform (DCT) and Discrete Sin Transform (DST) to increase the robustness of extracted features against different types of degradation through exploit the sub-band decomposition characteristics of DWT and the energy compaction property of DCT and DST. This achieved by extracting features directly from speech signal contaminated with features extracted from Discrete transformed signals to create a hybrid feature vectors. The enhancement techniques such as Spectral Subtraction (SS), Winer Filter (WF) and Adaptive Wiener Filter as a preprocessing stage to eliminate the effect of the interference on the ASI performance. In the second model, we investigate the using of Deep Learning (DL) based Convolutional Neural Network (CNN) with the spectrogram and the Radon transform to increase the robustness of the ASI against the interference effects. One of this paper goals is making comparison between two models and reach to more robust ASI system against severe interference. The experimental results indicates that the two proposed models articulate satisfying results, the model-based CNN consuming time is less than that based MLP which requires large training epochs and using many techniques for feature extraction and enhancing performance.

show abstract

“…Instead of using MFCC features, CI users performance in speaker identification is quantified by extracting electrodograms from speech signal using CI processor. A normal hearing person can perform efficient speaker recognition when the individual-speaker features are as widely separated as possible [9,10,11,12,13]. Based on this feature extraction mechanism, MFCC have been widely used in automated speaker recognition system as an attempt to mimic the speaker recognition capabilities of humans.…”

Section: Introductionmentioning

confidence: 99%

Quantifying Cochlear Implant Users’ Ability for Speaker Identification Using CI Auditory Stimuli

2019

Self Cite

View full text Add to dashboard Cite

Speaker recognition is a biometric modality that uses underlying speech information to determine the identity of the speaker. Speaker Identification (SID) under noisy conditions is one of the challenging topics in the field of speech processing, specifically when it comes to individuals with cochlear implants (CI). This study analyzes and quantifies the ability of CI-users to perform speaker identification based on direct electric auditory stimuli. CI users employ a limited number of frequency bands (8 ∼ 22) and use electrodes to directly stimulate the Basilar Membrane/Cochlear in order to recognize the speech signal. The sparsity of electric stimulation within the CI frequency range is a prime reason for loss in human speech recognition, as well as SID performance. Therefore, it is assumed that CIusers might be unable to recognize and distinguish a speaker given dependent information such as formant frequencies, pitch etc. which are lost to un-simulated electrodes. To quantify this assumption, the input speech signal is processed using a CI Advanced Combined Encoder (ACE) signal processing strategy to construct the CI auditory electrodogram. The proposed study uses 50 speakers from each of three different databases for training the system using two different classifiers under quiet, and tested under both quiet and noisy conditions. The objective result shows that, the CI users can effectively identify a limited number of speakers. However, their performance decreases when more speakers are added in the system, as well as when noisy conditions are introduced. This information could therefore be used for improving CI-user signal processing techniques to improve human SID.

show abstract

A Robust Text Dependent Speaker Identification Using Neural Responses from the Model of the Auditory System

Cited by 4 publications

References 10 publications

Probabilistic Permutation Invariant Training for Speech Separation

Probabilistic Permutation Invariant Training for Speech Separation

Automatic Speaker Identification System based MLP Network and Deep Learning in the Presence of Severe Interference

Quantifying Cochlear Implant Users’ Ability for Speaker Identification Using CI Auditory Stimuli

Contact Info

Product

Resources

About