結合不同層級特徵於語音情緒辨識之研究

Combining Different Levels of Features for Emotion Recognition in Speech

李芝宇(Chih-Yu Lee)

指導教授 ： 張智星

國立清華大學/電機資訊學院/資訊工程學系/碩士(2011年)

https://doi.org/10.6843/NTHU.2011.00046

摘要

情緒辨識廣泛的被應用在很多領域中。正確的辨識情緒一直是這項研究的目的。我們認為不同階層的語音特徵可以提供不同的語音資訊，並且相信結合不同階層的語音特徵可以提高辨識率。從實驗結果可得知，將不同階層的語音特徵結合再一起，可以有效達到彌補各階層資訊不足的情況。我們提出幾種不同階層特徵之組合方式，並且在實驗中，我們證實了組合不同階層語音特徵確實可以提升辨識率。在實驗中我們採用德語情緒語料庫以及英語eNTERFACE情緒語料庫，前者有七種情緒類別，後者則有六種情緒類別。我們擷取的語音特徵值可分為兩種，第一類是以音框階層為基準擷取的語音特徵，包含能量、音高以及梅爾倒頻譜係數；第二類則是針對區段階層以及語句階層擷取，擷取的特徵則為low-level-descriptors (LLDs)。由實驗可知，相較於單一階層的語音特徵，結合多層特徵將能有效提升辨識率。

關鍵字

情緒辨識

並列摘要

Emotion recognition has been successfully applied in many fields. It is believed that features extracted from each timing-level can provide different information of the emotional speech signals and therefore can compensate one another. In order to achieve a promising recognition accuracy, several methods for combining features extracted from different timing-levels are proposed in this thesis, including likelihood combination, weighted likelihood combination, raw feature combination and partial raw feature combination. We extracted spectrum features and prosodic features for frame-level features, and low-level descriptors (LLDs) for segment-level features and utterance-level features. The Berlin Emotion Database and eNTERFACE emotional database are used in the experiments. Compared with conventional one or two timing-level features, the combination of three timing-level features shows higher recognition rate.

參考文獻

[1] Dan-Ning Jiang and Lian-Hong Cai, “Speech emotion classification with the combination of statistic features and temporal features”, (2004) IEEE ICME

[2] B. Schuller and Gerhard Rigoll, “Timing levels in segment-based speech emotion recognition”, (2006) Interspeech

[6] M. Chetouani, A. Mahdhaoui and F. Ringeval, “Time-scale feature extractions for emotional speech characterization”, (2009) Cognitive Computation, 194-201

[7] Yi-Lin Lin and Gang Wei, “Speech emotion recognition based on GMM and SVM”, Fourth Internation Conference on Machine Learning and Cybernetics, (2005) 18-21

[8] D. Ververidis ,C. Kotropoulos and Ioannis Pitas, “Automatic emotional speech classification”, (2004) 593-596