arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1009.0117cs.LG
收藏

基于特征选择的情感声学特征研究

Exploring Language-Independent Emotional Acoustic Features via Feature Selection

School of Computer Science, The University of Manchester(曼彻斯特大学计算机科学学院)

Arslan Shaukat, Ke Chen

摘要:

我们提出了一种新的特征选择策略来发现语言无关的声学特征,这些特征往往与语言,语言学和其他因素无关。实验结果表明,语言无关的特征子集发现产生的性能相媲美的各种情感语音语料库的全特征集。

英文摘要:

We propose a novel feature selection strategy to discover language-independent acoustic features that tend to be responsible for emotions regardless of languages, linguistics and other factors. Experimental results suggest that the language-independent feature subset discovered yields the performance comparable to the full feature set on various emotional speech corpora.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑