Spectral Features Based on Local Hu Moments of Gabor Spectrograms for Speech Emotion Recognition

Abstract

SUMMARYTo improve the recognition rate of the speech emotion, new spectral features based on local Hu moments of Gabor spectrograms are proposed, denoted by GSLHu-PCA. Firstly, the logarithmic energy spectrum of the emotional speech is computed. Secondly, the Gabor spectrograms are obtained by convoluting logarithmic energy spectrum with Gabor wavelet. Thirdly, Gabor local Hu moments(GLHu) spectrograms are obtained through block Hu strategy, then discrete cosine transform (DCT) is used to eliminate correlation among components of GLHu spectrograms. Fourthly, statistical features are extracted from cepstral coefficients of GLHu spectrograms, then all the statistical features form a feature vector. Finally, principal component analysis (PCA) is used to reduce redundancy of features. The experimental results on EmoDB and ABC databases validate the effectiveness of GSLHu-PCA.

Parallelized Convolutional Recurrent Neural Network With Spectral Features for Speech Emotion Recognition

“…Effects of the experiment are conspicuous. On the ABC dataset, our method also outperforms [33], [37], [38], [40] in term of WA. And we report an UA of 57.59% on the ABC dataset, on which outperforms all the four compared works, i.e., 56.1% by [37], 55.5% by [33], 56.11% by [38], 52.26% by [40].…”

“…On the ABC dataset, our method also outperforms [33], [37], [38], [40] in term of WA. And we report an UA of 57.59% on the ABC dataset, on which outperforms all the four compared works, i.e., 56.1% by [37], 55.5% by [33], 56.11% by [38], 52.26% by [40]. On the EMO-DB dataset, our method also clearly outperforms all the four compared works.…”

A Novel Heterogeneous Parallel Convolution Bi-LSTM for Speech Emotion Recognition

“…1 to 3 summarizes the improvement of the performances of HPCB in terms of UAR to the related peer methods on databases CASIA, EMODB, and SAVEE. Among them, literatures [7], [29][30], [32] used the research results of previous researchers as the baseline, while literature [21] was originally proposed in the research of automatic speech recognition. When researchers in literature [34][35][36][37] applied it to speech emotion recognition, the database used was also inconsistent with the database used in this study.…”

A Novel Heterogeneous Parallel Convolution Bi-LSTM for Speech Emotion Recognition

“…Tables 1-3 summarize the performance improvements of HPCB in terms of UAR with respect to the related peer methods on the databases CASIA, EMODB, and SAVEE. Among them, the authors of [9,[41][42][43][44] used the research results of previous researchers as the baseline, while the study of [45] was originally proposed in the research of automatic speech recognition. When researchers in [46][47][48][49] applied it to speech emotion recognition, the database used was also inconsistent with the database used in this study.…”

“…GA-BEL [41] 38.55 38.55 HuWSF [42] 43.50 43.50 RDBN [44] 48.50 48.50 PCRN [9] 58.25 58.25 Bi-LSTM [46] / 75.00 Bi-GRU [47] / 72.50 CNN [48] / 76.67 CLDNN [45] / 61.67 CapsNet [49] / 63.33 HPCB (Ours) / 79.67…”