SVM Based Speaker Verification using a GMM Supervector Kernel and NAP Variability Compensation
By William M. Campbell, Douglas E. Sturim, Douglas A. Reynolds, Alex Solomonoff. 2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings, 2006
682 citations recorded by the source corpus.
Subfields: Artificial Intelligence, Signal Processing
View Paper: https://www.semanticscholar.org/paper/a9bcb7ff4b1caf20f3c23adee5fa43d92074840d
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
- Patrick Kenny
- Najim Dehak
- Réda Dehak
- Pierre Dumouchel
- Ahmad Salman
- Pierre Ouellet
- Daniel Povey
- Bing Jiang
In Signal Processing
- Patrick Kenny
- Najim Dehak
- Réda Dehak
- Pierre Dumouchel
- Ahmad Salman
- Pierre Ouellet
- Bing Jiang
- Daniel Povey
Citations in the corpus
- A review of depression and suicide risk assessment using speech analysis (2015, methodology): “Nuisance mitigating Many nuisance mitigation techniques exist from speech and speaker recognition that can help improve robustness against non-ideal acoustic conditions and confounding forms of variability, such as Nuisance Attribute Projection (NAP), i-vectors and Joint Factor Analysis (JFA) (Campb”
- Support vector machines versus fast scoring in the low-dimensional total variability space for speaker verification (2009, methodology): “The third and last approach is the Nuisance Attribute Projection (NAP), presented in [5].”
- Front-End Factor Analysis for Speaker Verification (2011, methodology): “Thelastapproachisthenuisanceattributeprojection (NAP) presented in [ 5 ].”
- Learning Speaker-Specific Characteristics With a Deep Neural Architecture (2011, methodology): “NAP is a compensation procedure for environmental variability used in SVM framework [111], [112].”
- Multi-modal Egocentric Activity Recognition using Audio-Visual Features (2018, methodology): “The mean vectors of the components of the resulting class-dependent GMM are then concatenated to form a ‘supervector’ [27].”
- Deep Speaker: an End-to-End Neural Speaker Embedding System (2017, background): “Though many algorithms have pushed the state-of-the-art over the past couple years [1][2][3][4][5][6][7], speaker recognition is still a challenging task.”
- What Does the Speaker Embedding Encode? (2017, positive): “Super-vector was proposed and achieved competitive results when combined with a SVM classifier [3, 4].”