Support vector machines for speaker and language recognition

By William M. Campbell, Joseph P. Campbell Jr., Douglas A. Reynolds, Elliot Singer, Pedro A. Torres-Carrasquillo. Computer Speech and Language, 2006

600 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Signal Processing

View Paper: https://www.semanticscholar.org/paper/30eb6b47b4774fff7d1999b1059148800de30cfd

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Li Deng
2. Haizhou Li
3. Yoshua Bengio
4. Xiaodong He
5. Jianfeng Gao
6. Dong Yu
7. L. V. D. van der Maaten
8. Najim Dehak

In Signal Processing

1. Haizhou Li
2. Li Deng
3. Najim Dehak
4. Douglas A. Reynolds
5. Xiaodong He
6. Kong Aik LEE
7. Bin Ma
8. Georg Heigold

Citations in the corpus

• Language Recognition via i-vectors and Dimensionality Reduction (2011, background): “Gaussian mixture models (GMM) [1, 2] and support vector machines (SVM) [3] have been the classifiers of choice over recent years for acoustic modeling, consistently outperforming their high-level counterparts.”
• Spoken Language Recognition: From Fundamentals to Practice (2013, background): “For languages that they are not familiar with, human listeners can often make subjective judgments with reference to the languages they know, e.g., it sounds like Arabic, it is tonal like Mandarin or Vietnamese, or it has a stress pattern like German or English.”
• Learning Discriminative Fisher Kernels (2011, methodology): “A traditional approach to the classification of a structured object Xn = {xn1, . . . ,xnT } with xnt ∈ Rd is to train a generative model for each of the classes y ∈ {1, 2, . . . , C}, and to use the resulting generative models in a Bayes classifier through p(yn = c|Xn) ∝ p(Xn|yn = c)p(yn = c).”
• An Overview of Speaker Identification: Accuracy and Robustness Issues (2011, methodology): “The 1 5aij, x S ij6i51 N , d j 2 are obtained from the SVM optimization algorithm [31], [32], [33] given the t i j ideal outputs (ti j 5 1 for class 0 and ti j 5 21 for class 1), the class 0 (target j) and class 1 (all other) training data, and the kernel function K 1 x , y S2 .”
• Optimization Algorithms and Applications for Speech and Language Processing (2013, background): “In the early 1990s, Brown et al. [32] of IBM conducted the pioneering work in statistical modeling approaches, establishing a range of IBM models, known prosaically as “Model 1” through “Model 5.””
• A review of biometric technology along with trends and prospects (2014, background): “Table 4: Summary of behavioral modalities: Modality Feature Set Recognition Techniques Datasets Voice Spectrum, Glottal pulse features, pitch, energy, duration, rhythm, temporal features, phones, idiolects, semantics, accent, pronunciation, LPCC [204], MFCC [205], VQ [206-208], HMM [209], GMM [210, ”