The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
The reverberant speech segregation is a basic problem in speech enhancement and automatic speech recognition. Based on the deep neural networks (DNN), a novel binaural speech segregation method is proposed. The binaural feature is extracted and used as the cue to train a DNN with a ideal parameter mask. The trained DNN is used to distinguish the target speech and noise, and output the estimated parameter...
Within the framework of computational auditory scene analysis (CASA), a parameter masks estimator based on deep neural networks (DNN) is proposed for automatic speech recognition (ASR) in noisy environments. This paper addresses the robustness in binaural machine speech recognition by speech energy estimation using DNN. An ideal parameter mask (IPM) is introduced as the goal of the DNN estimator,...
Gammatone filterbanks are widely used in computational auditory models for modeling the peripheral filtering function of the cochlea. However, the high computational complexity and time consumption limits its usage in portable acoustic applications. To address this issue, a realtime and efficient digital implementation of Gammatone filterbank is proposed. The decomposed signal can be resynthesized...
Robust speech enhancement is a challenge task, especially in noisy environments. The deep neural network has shown good performance on binaural speech enhancement with various speakers at a same distance. As binaural cues are based on the locations of sound sources, this paper analyze the performance of binaural deep neural network with different distances. The theoretical derivation and experiment...
Speech signal degradation in real environments mainly results from room reverberation and concurrent noise. While human listening is robust in complex auditory scenes, current speech segregation algorithms do not perform well in noisy and reverberant environments. We treat the binaural segregation problem as binary classification, and employ deep neural networks (DNNs) for the classification task...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.