Learning Speaker-Specific Characteristics With a Deep Neural Architecture

By Ahmad Salman. 2011

118 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Signal Processing

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

In Signal Processing

Citations in the corpus

Deep learning in neural networks: An overview (2014, background): “UL can also serve to extract invariant features from different data items (e.g., Becker, 1991) through coupled NNs observing two different inputs (Schmidhuber & Prelinger, 1992), also called Siamese NNs (e.g., Bromley et al., 1993; Chen & Salman, 2011; Hadsell, Chopra, & LeCun, 2006; Taylor, Spiro, ”
Deep neural network-based speaker embeddings for end-to-end speaker verification (2016, background): “Earlier studies in [21, 22, 23] presented a similar architecture that trains a DNN on a speaker comparison task and produces frame-level features that capture speaker characteristics.”
Selection of Proper Neural Network Sizes and Architectures—A Comparative Study (2012, background): “works [67], [68] which consist of multiple hidden layers connected in MLP architecture.”
Speaker Recognition for Multi-speaker Conversations Using X-vectors (2019, methodology): “Early work using discriminatively trained neural networks to capture speaker characteristics focused on extracting frame-level features to be used as input to Gaussian speaker models [6, 7].”
Time series forecasting using a deep belief network with restricted Boltzmann machines (2014, methodology): “DBN has been successfully applied to dimensionality reduction (image compression) [10], digit recognition [11], acoustic representation [12], and it is still developed ongoing [13].”
Deep Neural Network Embeddings for Text-Independent Speaker Verification (2017, background): “In early systems, neural networks are trained to classify training speakers [16, 17] or in Siamese architectures to separate same-speaker and differentspeaker pairs [18, 19, 20].”
X-Vectors: Robust DNN Embeddings for Speaker Recognition (2018, methodology): “In early systems, neural networks are trained to separate speakers, and frame-level representations are extracted from the network and used as features for Gaussian speaker models [14, 15, 16].”
Speaker identification and clustering using convolutional neural networks (2016, background): “trained a siamese deep network to directly compare two voice segments discriminatively [18].”