Attention-Based Dense LSTM for Speech Emotion Recognition

Yue XIE  Ruiyu LIANG  Zhenlin LIANG  Li ZHAO  

IEICE TRANSACTIONS on Information and Systems   Vol.E102-D    No.7    pp.1426-1429
Publication Date: 2019/07/01
Publicized: 2019/04/17
Online ISSN: 1745-1361
DOI: 10.1587/transinf.2019EDL8019
Type of Manuscript: LETTER
Category: Pattern Recognition
attention mechanism,  speech emotion recognition,  dense connections,  LSTM,  

Full Text: FreePDF(322.6KB)

Despite the widespread use of deep learning for speech emotion recognition, they are severely restricted due to the information loss in the high layer of deep neural networks, as well as the degradation problem. In order to efficiently utilize information and solve degradation, attention-based dense long short-term memory (LSTM) is proposed for speech emotion recognition. LSTM networks with the ability to process time series such as speech are constructed into which attention-based dense connections are introduced. That means the weight coefficients are added to skip-connections of each layer to distinguish the difference of the emotional information between layers and avoid the interference of redundant information from the bottom layer to the effective information from the top layer. The experiments demonstrate that proposed method improves the recognition performance by 12% and 7% on eNTERFACE and IEMOCAP corpus respectively.

open access publishing via