posted by user: bairdali || 1794 views || tracked by 2 users: [display]

MuSe 2021 2021 : The 2nd International Multimodal Sentiment Analysis in Real-life Media Challenge and Workshop @ ACM Multimedia 2021, October 2021, Chengdu, China


When Oct 20, 2021 - Oct 24, 2021
Where Chengdu, China
Submission Deadline Aug 5, 2021
Notification Due Aug 26, 2021
Final Version Due Sep 2, 2021
Categories    affective computing   machine learning   sentiment analysis

Call For Papers

MuSe 2021 focuses on multimodal sentiment/ emotion recognition of user-generated content and in stress-induced situations: This year, we focus on audio-visual and textual, as well as biological signals (inc. heart rate and galvanic skin response).

1. Multimodal Continuous Emotions in-the-Wild Sub-challenge (MuSe-Wilder)
Predicting the level of emotional dimensions (arousal, valence) in a time-continuous manner from audio-visual recordings.
2. Multimodal Sentiment Sub-challenge (MuSe-Sent)
Predicting advanced intensity classes of emotions based on valence and arousal for segments of audio-visual recordings.
3. Multimodal Emotional Stress Sub-challenge (MuSe-Stress)
Predicting the level of emotional arousal and valence in a time-continuous manner from audio-visual recordings.
4. Multimodal Physiological-Arousal Sub-challenge (MuSe-Physio)
Predicting the level of psycho-physiological arousal from a) human annotations fused with b) galvanic skin response (also known as Electrodermal Activity (EDA)) signals from the stressed people as a regression task. Audio-visual recordings, as well as other biological signals (heart rate and respiration), are offered for modelling.

== Datasets ==
- MuSe-CaR: The MuSe-Wilder and MuSe-Sent sub-challenges use the MuSe-CAR database. It is a large (~40h), a multimodal dataset that has been gathered in-the-wild with the intention of further understanding Multimodal Sentiment Analysis, e.g., the emotional engagement that takes place during product reviews (i.e., automobile reviews) where sentiment is linked to a topic or entity.
- Ulm-TSST: For MuSe-Physio and MuSe-Stress, the novel Ulm-TSST database provides a multimodal dataset of emotion and mental well-being. It includes biological recordings, such as Electrocardiogram (ECG), Electrodermal Activity (EDA), Respiration, and Heart Rate (BPM) as well as continuous arousal and valence annotations. With 105 participants aged between 18 and 39 years, a total of 10 hours were accumulated.
The test set labels are held back by the organizers for evaluation.

== Features ==
Besides the raw data, video (Facial Action Units, VGGFace, Xception), audio (VGGish, EGeMAPS, DeepSpectrum), and text (BERT) features are provided.

== How to Participate ==
Instructions are available at
Data, features and baseline model (code, weights) are available and described in a preliminary paper.

== Organisers ==
Björn W. Schuller (Imperial College London/ audEERING, UK)
Eva-Maria Meßner (University of Ulm, GER)
Erik Cambria (Nanyang Technological University/ SenticNet, SG)
Guoying Zhao (University of Oulu)
Lukas Stappen (University of Augsburg/ BMW Group, GER)

Related Resources

ALLEGET 2022   2nd International Workshop on Artificial Intelligence and Machine Learning for Emerging Topics
ICML 2022   39th International Conference on Machine Learning
MUWS 2022   The 1st International Workshop on Multimodal Understanding for the Web and Social Media, co-located with The WebConf (WWW)
IEEE COINS 2022   IEEE COINS 2022: Hybrid (3 days on-site | 2 days virtual)
SALLD 2022   2nd Workshop on Sentiment Analysis & Linguistic Linked Data
JCRAI 2022-Ei Compendex & Scopus 2022   2022 International Joint Conference on Robotics and Artificial Intelligence (JCRAI 2022)
Analysis and Mining of Social Media Data 2022   Analysis and Mining of Social Media Data
MLDM 2022   18th International Conference on Machine Learning and Data Mining
MARS 2022   Workshop on Models for Formal Analysis of Real Systems
FAIML 2022   2022 International Conference on Frontiers of Artificial Intelligence and Machine Learning (FAIML 2022)