M
Meet H. Soni
Researcher at Dhirubhai Ambani Institute of Information and Communication Technology
Publications - 22
Citations - 476
Meet H. Soni is an academic researcher from Dhirubhai Ambani Institute of Information and Communication Technology. The author has contributed to research in topics: Autoencoder & Noise. The author has an hindex of 7, co-authored 19 publications receiving 339 citations. Previous affiliations of Meet H. Soni include Tata Consultancy Services & Harvard University.
Papers
More filters
Proceedings ArticleDOI
Time-Frequency Masking-Based Speech Enhancement Using Generative Adversarial Network
TL;DR: The proposed system significantly improves over a recent GAN-based speech enhancement system in improving speech quality, while maintaining a better trade-off between less speech distortion and more effective removal of background interferences present in the noisy mixture.
Proceedings ArticleDOI
Novel Variable Length Teager Energy Separation Based Instantaneous Frequency Features for Replay Detection.
TL;DR: A novel replay detector based on Variable length Teager Energy OperatorEnergy Separation Algorithm-Instantaneous Frequency Cosine Coefficients (VESA-IFCC) for the ASV spoof 2017 challenge is proposed and the performance of the proposed feature set is compared with the features developed for detecting synthetic and voice converted speech.
Proceedings ArticleDOI
Novel TEO-based Gammatone features for environmental sound classification
TL;DR: Modified Gammatone filterbank with Teager Energy Operator (TEO) with two classifiers, namely, Gaussian Mixture Model (GMM) using cepstral features, and Convolutional Neural Network (CNN) using spectral features are used for environmental sound classification (ESC) task.
Proceedings ArticleDOI
Novel deep autoencoder features for non-intrusive speech quality assessment
Meet H. Soni,Hemant A. Patil +1 more
TL;DR: Quantification of the experimental results suggests that proposed metric gives more accurate and correlated scores than an existing benchmark for objective, non-intrusive quality assessment metric ITU-T P.563 standard.
Proceedings ArticleDOI
Time-Frequency Mask-based Speech Enhancement using Convolutional Generative Adversarial Network
TL;DR: This paper proposes to use a Convolutional Neural Network-based Generative Adversarial Network (GAN) for inherent mask estimation, an alternative to the other Maximum Likelihood (ML) optimization-based architectures and shows the need for supervised T-F mask estimation for effective noise suppression.