MSER: Multimodal speech emotion recognition using cross-attention with deep fusion

计算机科学 稳健性(进化) 语音识别 判别式 人工智能 编码器 特征(语言学) 融合机制 模式识别(心理学) 融合 基因 操作系统 脂质双层融合 哲学 生物化学 化学 语言学
作者
Mustaqeem Khan,Wail Gueaieb,Abdulmotaleb El Saddik,Soonil Kwon
出处
期刊:Expert Systems With Applications [Elsevier]
卷期号:245: 122946-122946 被引量:23
标识
DOI:10.1016/j.eswa.2023.122946
摘要

In human-computer interaction (HCI) and especially speech signal processing, emotion recognition is one of the most important and challenging tasks due to multi-modality and limited data availability. Nowadays, an intelligent system is required for real-world applications to efficiently process and understand the speaker's emotional state and to enhance the analytical abilities to assist communication by a human-machine interface (HMI). Designing a reliable and robust Multimodal Speech Emotion Recognition (MSER) to efficiently recognize emotions through multi-modality such as speech and text is necessary. This paper proposes a novel MSER model with a deep feature fusion technique using a multi-headed cross-attention mechanism. The proposed model utilizes audio and text cues to predict the emotion label accordingly. Our proposed model processes the raw speech signal and text by CNN and feeds to corresponding encoders for discriminative and semantic feature extractions. The cross-attention mechanism is applied to both features to enhance the interaction between text and audio cues by crossway to extract the most relevant information for emotion recognition. Finally, combining the region-wise weights from both encoders enables interaction among different layers and paths by the proposed deep feature fusion scheme. The authors evaluate the proposed system using the IEMOCAP and MELD datasets and conduct extensive experiments that obtain state-of-the-art (SOTA) results and show a 4.5% improved recognition rate, respectively. Our model secured a significant improvement over SOTA methods, which shows the robustness and effectiveness of the proposed MSER model.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
更新
大幅提高文件上传限制,最高150M (2024-4-1)

科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
aabb完成签到,获得积分10
刚刚
1秒前
赵zhao完成签到,获得积分10
1秒前
3秒前
椒盐丸子完成签到,获得积分10
4秒前
请叫我鬼才完成签到,获得积分10
6秒前
丘比特应助叮咚采纳,获得10
7秒前
7秒前
yu完成签到 ,获得积分10
7秒前
8秒前
Monet发布了新的文献求助10
8秒前
9秒前
青木完成签到,获得积分10
10秒前
10秒前
拾柒完成签到,获得积分10
10秒前
11秒前
农学小王完成签到 ,获得积分10
11秒前
khr完成签到,获得积分20
11秒前
yoyo发布了新的文献求助10
12秒前
领导范儿应助wbbb采纳,获得10
12秒前
14秒前
duoduo发布了新的文献求助10
14秒前
大模型应助Monet采纳,获得10
15秒前
16秒前
16秒前
科研通AI2S应助科研通管家采纳,获得10
16秒前
在水一方应助科研通管家采纳,获得10
16秒前
小二郎应助科研通管家采纳,获得30
16秒前
16秒前
xiaoliu应助科研通管家采纳,获得10
16秒前
淡淡芷天应助科研通管家采纳,获得10
16秒前
16秒前
丘比特应助科研通管家采纳,获得10
16秒前
劲秉应助科研通管家采纳,获得30
16秒前
orixero应助科研通管家采纳,获得10
17秒前
科研通AI2S应助科研通管家采纳,获得10
17秒前
劲秉应助科研通管家采纳,获得30
17秒前
英俊的铭应助科研通管家采纳,获得10
17秒前
英俊的铭应助科研通管家采纳,获得10
17秒前
科目三应助科研通管家采纳,获得10
17秒前
高分求助中
Medicina di laboratorio. Logica e patologia clinica 600
Sarcolestes leedsi Lydekker, an ankylosaurian dinosaur from the Middle Jurassic of England 500
《关于整治突出dupin问题的实施意见》(厅字〔2019〕52号) 500
Language injustice and social equity in EMI policies in China 500
mTOR signalling in RPGR-associated Retinitis Pigmentosa 500
A new species of Velataspis (Hemiptera Coccoidea Diaspididae) from tea in Assam 500
Geochemistry, 2nd Edition 地球化学经典教科书第二版 401
热门求助领域 (近24小时)
化学 医学 生物 材料科学 工程类 有机化学 生物化学 物理 内科学 纳米技术 计算机科学 化学工程 复合材料 基因 遗传学 催化作用 物理化学 免疫学 量子力学 细胞生物学
热门帖子
关注 科研通微信公众号,转发送积分 3214262
求助须知:如何正确求助?哪些是违规求助? 2862914
关于积分的说明 8135930
捐赠科研通 2529163
什么是DOI,文献DOI怎么找? 1363278
科研通“疑难数据库(出版商)”最低求助积分说明 643775
邀请新用户注册赠送积分活动 616269