Variational Recurrent Neural Networks for Speech Separation
نویسندگان
چکیده
We present a new stochastic learning machine for speech separation based on the variational recurrent neural network (VRNN). This VRNN is constructed from the perspectives of generative stochastic network and variational auto-encoder. The idea is to faithfully characterize the randomness of hidden state of a recurrent neural network through variational learning. The neural parameters under this latent variable model are estimated by maximizing the variational lower bound of log marginal likelihood. An inference network driven by the variational distribution is trained from a set of mixed signals and the associated source targets. A novel supervised VRNN is developed for speech separation. The proposed VRNN provides a stochastic point of view which accommodates the uncertainty in hidden states and facilitates the analysis of model construction. The masking function is further employed in network outputs for speech separation. The benefit of using VRNN is demonstrated by the experiments on monaural speech separation.
منابع مشابه
Sequential Neural Models with Stochastic Layers
How can we efficiently propagate uncertainty in a latent state representation with recurrent neural networks? This paper introduces stochastic recurrent neural networks which glue a deterministic recurrent neural network and a state space model together to form a stochastic and sequential neural generative model. The clear separation of deterministic and stochastic layers allows a structured va...
متن کاملAdaptive Speech Separation Using Hybrid Approach
A hybrid iterative learning algorithm for recurrent neural networks based on higher-order statistics to blind signal separation is presented in this paper. Fourth-order statistics are used as the separation criterion to train an RNN to perform the separation. Some simulation results for both artificially convoluted audio signals and real recordings demonstrate that the proposed approach is prom...
متن کاملSpeech Emotion Recognition Using Scalogram Based Deep Structure
Speech Emotion Recognition (SER) is an important part of speech-based Human-Computer Interface (HCI) applications. Previous SER methods rely on the extraction of features and training an appropriate classifier. However, most of those features can be affected by emotionally irrelevant factors such as gender, speaking styles and environment. Here, an SER method has been proposed based on a concat...
متن کاملBinaural Speech Separation Using Recurrent Timing Neural Networks for Joint F0-Localisation Estimation
A speech separation system is described in which sources are represented in a joint interaural time difference-fundamental frequency (ITD-F0) cue space. Traditionally, recurrent timing neural networks (RTNNs) have been used only to extract periodicity information; in this study, this type of network is extended in two ways. Firstly, a coincidence detector layer is introduced, each node of which...
متن کامل