A Semisupervised Approach for Language Identification based on Ladder Networks
This work addresses language identification for speech processing, but it is incremental as it adapts an existing semisupervised method to a specific domain.
The authors tackled the problem of language identification using both labeled and unlabeled speech data, proposing a neural network based on Ladder Networks that also handles out-of-set languages, and achieved enhanced performance on the NIST 2015 dataset.
In this study we address the problem of training a neuralnetwork for language identification using both labeled and unlabeled speech samples in the form of i-vectors. We propose a neural network architecture that can also handle out-of-set languages. We utilize a modified version of the recently proposed Ladder Network semisupervised training procedure that optimizes the reconstruction costs of a stack of denoising autoencoders. We show that this approach can be successfully applied to the case where the training dataset is composed of both labeled and unlabeled acoustic data. The results show enhanced language identification on the NIST 2015 language identification dataset.