Sivanand Achanta

TL;DR: A variant of DTW based algorithm referred to as non-segmental DTW (NS-DTW) is used, with a computational upper bound of O (mn) and analyzed the performance of QbE-STD with Gaussian posteriorgrams obtained from spectral and temporal features of the speech signal, showing that frequency domain linear prediction cepstral coefficients can be used as an alternative to traditional spectral parameters.

TL;DR: The experimental results on ASVspoof 2017 dataset reveal that, SFF based representation is very effective in detecting replay attacks and the score level fusion of back end classifiers further improved the performance of the system which indicates that both classifiers capture complimentary information.

TL;DR: In this paper, deep neural networks are investigated for language identification in Indian languages and an attention mechanism based DNN architecture is proposed for utterance level classification there by efficiently making use of the context.

TL;DR: It is shown that clockwork RNN is equivalent to an Elman RNN with a particular form of LI, and this perspective enables us to understand the reason why a simple ElmanRNN with LI units performs well on sequential tasks.

TL;DR: Deep Elman RNNs are better suited for acoustic modeling in SPSS when compared to DNNs and perform competitively to deep SLSTMs, GRUs and LSTM, and context representation learning using ElmanRNNs improves neural network acoustic models for S PSS.