Uncertain LDA: Including Observation Uncertainties in Discriminative Transforms

Saeidi, Rahim; Astudillo, Ramón Fernandez; Kolossa, Dorothea

doi:10.1109/tpami.2015.2481420

Cited by 31 publications

(32 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…where W is the projection matrix. By solving the above equation using generalized eigenvalue decomposition, uncertaintyaware channel compensation can be applied to i-vectors [29].…”

Section: B Digit Dependent Uncertainty and Channel Compensationmentioning

confidence: 99%

Speaker Recognition With Random Digit Strings Using Uncertainty Normalized HMM-Based i-Vectors

Maghsoodi

Sameti

Zeinali

et al. 2019

IEEE/ACM Trans. Audio Speech Lang. Process.

View full text Add to dashboard Cite

In this paper, we combine Hidden Markov Models (HMMs) with i-vector extractors to address the problem of textdependent speaker recognition with random digit strings. We employ digit-specific HMMs to segment the utterances into digits, to perform frame alignment to HMM states and to extract Baum-Welch statistics. By making use of the natural partition of input features into digits, we train digit-specific i-vector extractors on top of each HMM and we extract well-localized i-vectors, each modelling merely the phonetic content corresponding to a single digit. We then examine ways to perform channel and uncertainty compensation, and we propose a novel method for using the uncertainty in the i-vector estimates. The experiments on RSR2015 part III show that the proposed method attains 1.52% and 1.77% Equal Error Rate (EER) for male and female respectively, outperforming state-of-the-art methods such as xvectors, trained on vast amounts of data. Furthermore, these results are attained by a single system trained entirely on RSR2015, and by a simple score-normalized cosine distance. Moreover, we show that the omission of channel compensation yields only a minor degradation in performance, meaning that the system attains state-of-the-art results even without recordings from multiple handsets per speaker for training or enrolment. Similar conclusions are drawn from our experiments on the RedDots corpus, where the same method is evaluated on phrases. Finally, we report results with bottleneck features and show that further improvement is attained when fusing them with spectral features.

show abstract

“…where W is the projection matrix. By solving the above equation using generalized eigenvalue decomposition, uncertaintyaware channel compensation can be applied to i-vectors [29].…”

Section: B Digit Dependent Uncertainty and Channel Compensationmentioning

confidence: 99%

Speaker Recognition With Random Digit Strings Using Uncertainty Normalized HMM-Based i-Vectors

Maghsoodi

Sameti

Zeinali

et al. 2019

IEEE/ACM Trans. Audio Speech Lang. Process.

View full text Add to dashboard Cite

show abstract

“…us, we obtain λ � f a k+1 . (23) e solution to the problem is to maximize the value of λ. Multiply a T j S − 1 t (j � 1, .…”

Section: Solution To the Projection Matrix A Optmentioning

confidence: 99%

“…At present, PCA and LDA have a lot of applications in image processing, voice processing, communication, network, and others. Many researchers [9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27] have done extensibility research based on LDA and PCA methods and have made some progress. But there are some shortcomings in the use of PCA and LDA.…”

Section: Introductionmentioning

confidence: 99%

“…Meanwhile, face recognition is also an important way to verify the effectiveness of pattern recognition methods. e LDA-based methods are widely used in face recognition [23][24][25][26][27][30][31][32][33][34][35][36][37][38][39][40][41][42]. e research on LDA can be traced back to a classic paper [3] by Fisher in 1936. e basic idea is to choose the vector that makes the Fisher criterion function as max as optimal projection vector, so that the sample can achieve the maximum between-class scatter and the smallest within-class scatter after being projected in this direction.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Similar Distribution Discriminant Analysis with Orthogonal and Nearly Statistically Uncorrelated Characteristics

Guo

Zhang

2019

Mathematical Problems in Engineering

View full text Add to dashboard Cite

It is very difficult to process and analyze high-dimensional data directly. Therefore, it is necessary to learn a potential subspace of high-dimensional data through excellent dimensionality reduction algorithms to preserve the intrinsic structure of high-dimensional data and abandon the less useful information. Principal component analysis (PCA) and linear discriminant analysis (LDA) are two popular dimensionality reduction methods for high-dimensional sensor data preprocessing. LDA contains two basic methods, namely, classic linear discriminant analysis and FS linear discriminant analysis. In this paper, a new method, called similar distribution discriminant analysis (SDDA), is proposed based on the similarity of samples’ distribution. Furthermore, the method of solving the optimal discriminant vector is given. These discriminant vectors are orthogonal and nearly statistically uncorrelated. The disadvantages of PCA and LDA are overcome, and the extracted features are more effective by using SDDA. The recognition performance of SDDA exceeds PCA and LDA largely. Some experiments on the Yale face database, FERET face database, and UCI multiple features dataset demonstrate that the proposed method is effective. The results reveal that SDDA obtains better performance than comparison dimensionality reduction methods.

show abstract

“…Utterances with higher quality provide more reliable information for the model than those with lower quality. Thus, the contribution of each utterance in the recognition process should be controlled by a quality measure [27]. In addition to the typical quality measures such as signal-to-noise ratio (SNR), duration and F0 deviations, recently, uncertainty of the utterance modeling has been incorporated into the recognition process [16].…”

Section: Proposed I-vector Quality Measurementioning

confidence: 99%

Incorporating uncertainty as a Quality Measure in I-Vector Based Language Recognition

Poorjam¹,

Saeidi²,

Kinnunen³

et al. 2016

The Speaker and Language Recognition Workshop (Odyssey 2016)

Self Cite

View full text Add to dashboard Cite

State-of-the-art language recognition systems involve modeling utterances with the i-vectors. However, the uncertainty of the i-vector extraction process represented by the i-vector posterior covariance is affected by various factors such as channel mismatch, background noise, incomplete transformations and duration variability. In this paper, we propose a new quality measure based on the i-vector posterior covariance and incorporate it into the recognition process to improve the recognition accuracy. The experimental results with LRE15 database and various duration conditions show a 2.9% relative improvement in terms of average performance cost as a result of incorporating the proposed quality measure in language recognition systems. I-vector based language recognitionIn this section, we describe the main components of an ivector/PLDA-based language recognition system. The i-vector frameworkAn i-vector is a low-dimensional feature vector for representing utterances of arbitrary duration. We assume that each utterance possesses a speaker-and channel-dependent GMM mean supervector, M, in the form [5]:

show abstract

Uncertain LDA: Including Observation Uncertainties in Discriminative Transforms

Cited by 31 publications

References 53 publications

Speaker Recognition With Random Digit Strings Using Uncertainty Normalized HMM-Based i-Vectors

Speaker Recognition With Random Digit Strings Using Uncertainty Normalized HMM-Based i-Vectors

A Similar Distribution Discriminant Analysis with Orthogonal and Nearly Statistically Uncorrelated Characteristics

Incorporating uncertainty as a Quality Measure in I-Vector Based Language Recognition

Contact Info

Product

Resources

About