Conference paper

LOZANO Díez Alicia, PLCHOT Oldřich, MATĚJKA Pavel and GONZALEZ-RODRIGUEZ Joaquin. DNN Based Embeddings for Language Recognition. In: Proceedings of ICASSP 2018. Calgary: IEEE Signal Processing Society, 2018, pp. 5184-5188. ISBN 978-1-5386-4658-8.
Publication language:english
Original title:DNN Based Embeddings for Language Recognition
Title (cs):DNN Embeddings pro rozpoznávání jazyka
Proceedings:Proceedings of ICASSP 2018
Conference:IEEE International Conference on Acoustics, Speech and Signal Processing
Place:Calgary, CA
Publisher:IEEE Signal Processing Society
Embeddings, language recognition, LID, DNN
In this work, we present a language identification (LID) system based on embeddings. In our case, an embedding is a fixed-length vector (similar to i-vector) that represents the whole utterance, but unlike i-vector it is designed to contain mostly information relevant to the target task (LID). In order to obtain these embeddings, we train a deep neural network (DNN) with sequence summarization layer to classify languages. In particular, we trained a DNN based on bidirectional long short-term memory (BLSTM) recurrent neural network (RNN) layers, whose frame-by-frame outputs are summarized into mean and standard deviation statistics. After this pooling layer, we add two fully connected layers whose outputs correspond to embeddings. Finally, we add a softmax output layer and train the whole network with multi-class cross-entropy objective to discriminate between languages. We report our results on NIST LRE 2015 and we compare the performance of embeddings and corresponding i-vectors both modeled by Gaussian Linear Classifier (GLC). Using only embeddings resulted in comparable performance to i-vectors and by performing score-level fusion we achieved 7.3% relative improvement over the baseline.
   author = {Alicia D{\'{i}}ez Lozano and Old{\v{r}}ich Plchot
	and Pavel Mat{\v{e}}jka and Joaquin
   title = {DNN Based Embeddings for Language Recognition},
   pages = {5184--5188},
   booktitle = {Proceedings of ICASSP 2018},
   year = 2018,
   location = {Calgary, CA},
   publisher = {IEEE Signal Processing Society},
   ISBN = {978-1-5386-4658-8},
   doi = {10.1109/ICASSP.2018.8462403},
   language = {english},
   url = {}

Your IPv4 address:
Switch to https