arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2206.12469cs.SDcs.CLeess.AS
收藏

Burst2Vec:一种对抗性多任务方法,用于从声音爆发中预测情绪,年龄和起源

Burst2Vec: An Adversarial Multi-Task Approach for Predicting Emotion, Age, and Origin from Vocal Bursts

Atijit Anuchitanukul, Lucia Specia

摘要:

我们提出了Burst 2 Vec,我们的多任务学习方法来预测情绪,年龄和起源(即,本民族/语言)从声音爆发。Burst 2 Vec利用预先训练的语音表示来从原始波形中捕获声学信息,并通过对抗训练引入模型去偏置的概念。我们的模型使用预提取的特征实现了比基线相对30%的性能增益,并在ICML ExVo 2022多任务挑战赛的所有参与者中得分最高。

英文摘要:

We present Burst2Vec, our multi-task learning approach to predict emotion, age, and origin (i.e., native country/language) from vocal bursts. Burst2Vec utilises pre-trained speech representations to capture acoustic information from raw waveforms and incorporates the concept of model debiasing via adversarial training. Our models achieve a relative 30 % performance gain over baselines using pre-extracted features and score the highest amongst all participants in the ICML ExVo 2022 Multi-Task Challenge.

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。

↑