Global SNR Estimation of Speech Signals using Entropy and Uncertainty\n Estimates from Dropout Networks

2018/04/12 by Rohith Aralikatti, Aralikatti, Rohith, Dilip Kumar Margam +7
Computer Science · #Artificial Intelligence (cs.AI) #Audio and Speech Processing (eess.AS) #FOS: Computer and information sciences #FOS: Electrical engineering #Machine Learning (stat.ML) #Music and Audio Processing #Signal Processing (eess.SP) #Speech Recognition and Synthesis #Speech and Audio Processing #electronic engineering #information engineering

paper · pdf · doi:10.48550/arxiv.1804.04353

openalex publication_date 2018/04/12 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

This paper demonstrates two novel methods to estimate the global SNR of\nspeech signals. In both methods, Deep Neural Network-Hidden Markov Model\n(DNN-HMM) acoustic model used in speech recognition systems is leveraged for\nthe additional task of SNR estimation. In the first method, the entropy of the\nDNN-HMM output is computed. Recent work on bayesian deep learning has shown\nthat a DNN-HMM trained with dropout can be used to estimate model uncertainty\nby approximating it as a deep Gaussian process. In the second method, this\napproximation is used to obtain model uncertainty estimates. Noise specific\nregressors are used to predict the SNR from the entropy and model uncertainty.\nThe DNN-HMM is trained on GRID corpus and tested on different noise profiles\nfrom the DEMAND noise database at SNR levels ranging from -10 dB to 30 dB.\n

Related

• DNN-based uncertainty estimation for weighted DNN-HMM ASR
• Uncertainty weighting and propagation in DNN–HMM-based speech recognition
• Leveraging Deep Neural Network Activation Entropy to cope with Unseen\n Data in Speech Recognition
• Using Monte Carlo dropout for non-stationary noise reduction from speech
• Information Theoretic Analysis of DNN-HMM Acoustic Modeling