SEWA DB: A Rich Database for Audio-Visual Emotion and Sentiment Research in the Wild

Abstract

Natural human-computer interaction and audio-visual human behaviour sensing systems, which would achieve robust performance in-the-wild are more needed than ever as digital devices are increasingly becoming an indispensable part of our life. Accurately annotated real-world data are the crux in devising such systems. However, existing databases usually consider controlled settings, low demographic variability, and a single task. In this paper, we introduce the SEWA database of more than 2000 minutes of audio-visual data of 398 people coming from six cultures, 50% female, and uniformly spanning the age range of 18 to 65 years old. Subjects were recorded in two different contexts: while watching adverts and while discussing adverts in a video chat. The database includes rich annotations of the recordings in terms of facial landmarks, facial action units (FAU), various vocalisations, mirroring, and continuously valued valence, arousal, liking, agreement, and prototypic examples of (dis)liking. This database aims to be an extremely valuable resource for researchers in affective computing and automatic human sensing and is expected to push forward the research in human behaviour analysis, including cultural studies. Along with the database, we provide extensive baseline experiments for automatic FAU detection and automatic valence, arousal and (dis)liking intensity estimation.

References in corpus (1)

• Scikit-learn: Machine Learning in Python

Cited by in corpus (26)

• Dawn of the transformer era in speech emotion recognition: closing the valence gap
• Modeling emotion in complex stories: the Stanford Emotional Narratives Dataset
• On the use of Self-supervised Pre-trained Acoustic and Linguistic Features for Continuous Speech Emotion Recognition
• Multitask Learning and Multistage Fusion for Dimensional Audiovisual Emotion Recognition
• The Arousal video Game AnnotatIoN (AGAIN) Dataset
• Does Visual Self-Supervision Improve Learning of Speech Representations for Emotion Recognition?
• Predicting Player Engagement in Tom Clancy's The Division 2: A Multimodal Approach via Pixels and Gamepad Actions
• Deep Audio-Visual Learning: A Survey
• From the Lab to the Wild: Affect Modeling via Privileged Information
• Cross-Domain Facial Expression Recognition: A Unified Evaluation Benchmark and Adversarial Graph Learning
• End-to-End Label Uncertainty Modeling in Speech Emotion Recognition using Bayesian Neural Networks and Label Distribution Learning
• GameVibe: A Multimodal Affective Game Corpus
• Estimating the Uncertainty in Emotion Attributes using Deep Evidential Regression
• Remote smartphone-based speech collection: acceptance and barriers in individuals with major depressive disorder
• Multilingual Dyadic Interaction Corpus NoXi+J: Toward Understanding Asian-European Non-verbal Cultural Characteristics and their Influences on Engagement
• DEEMO: De-identity Multimodal Emotion Recognition and Reasoning
• SocialInteractionGAN: Multi-person Interaction Sequence Generation
• Adversarial-based neural networks for affect estimations in the wild
• Multimodal Continuous Emotion Recognition using Deep Multi-Task Learning with Correlation Loss
• An Estimation of Online Video User Engagement from Features of Continuous Emotions
• Continuous Emotion Recognition with Audio-visual Leader-follower Attentive Fusion
• M&M: Multimodal-Multitask Model Integrating Audiovisual Cues in Cognitive Load Assessment
• MuSe-Toolbox: The Multimodal Sentiment Analysis Continuous Annotation Fusion and Discrete Class Transformation Toolbox
• End-to-End Speech Emotion Recognition: Challenges of Real-Life Emergency Call Centers Data Recordings
• End-to-End Annotator Bias Approximation on Crowdsourced Single-Label Sentiment Analysis
• Translation, Sentiment and Voices: A Computational Model to Translate and Analyse Voices from Real-Time Video Calling