Induced Emotion Recognition with Personality Characteristics Challenge 2024

IERP 2024

个性化特征诱导情感识别挑战赛IERP 2024将会开放200人左右，涉及8种情感的多模态数据集，重点探究引入个性化特征的情感识别问题。欢迎大家参加！

Registration form：https://www.wjx.top/vm/OgVQpdA.aspx#

Challenge Email：IEPR2024@iscslp2024.com

Challenge Website：http://www.iscslp2024.com/emotionChallenges

About IERP 2024 Challenge

Induced Emotion Recognition with Personality Characteristics (IERP Challenge 2024) is part of the ISCSLP 2024 Competitions and Challenges track. While Emotion recognition has always been an active research topic in the field of artificial intelligence, existing emotion recognition mainly focuses on different technologies to improve the accuracy of discrete/dimensional emotion which limits its application. As well known, emotion has a direct link to personality, and people with different personality characteristics will have significant differences in emotional expression: neuroticism predicts higher negative emotions and lower positive affect, while conscientiousness, by contrast, is inversely associated with negative emotions. Therefore, this challenge aims to incorporate the factor of personality characteristics into the task of emotion recognition, focusing on the personalized information for future emotional applications.

Tracks

Track 1. Induced Emotion Recognition based on Audio modality:  The organizer will provide the speech features of each participant who watches the induced emotion videos. Participants are expected to recognize the scores of 8 types of emotion, sadness, relaxation, happiness, surprise, fear, anger, disgust, and neutrality, with the specified speech features provided by organizer. Score for each emotion ranges from 1 to 5, where 1 represents none and 5 represents the highest degree.

Track 2. Multi-model Induced Emotion Recognition:  The organizer will provide the speech features, text features and sound-muffling videos. Participants can freely design visual features and are expected to design effectiveness emotion classification method. The emotion type,  the score ranges,  and the speech features are same as Track 1. The  increment of this task is introducing text and visual information.

Data description

1. The database is recorded with a total of 193 participants, each with 16 videos.  Each video has self-rated scores for 8 types of emotions. The training set includes 133*16 samples, and the validation set and test set include 30*16 samples, respectively.

2. In track 1, only the speech features are provided. And in track 2, the speech features, text features and the original visual data are provided.

Important dates

18 May 2024: Registration for the challenge opens

28 May 2024: Release of the baseline system,Train and Eval data

23 Jun 2024: Release test data

9 Jul 2024: Submission deadline

12 Jul 2024: Announcement of Winners in all tracks

Organizers

Shan Liang,  Xi'an Jiaotong-Liverpool University, China
Xuefei Liu,  Institute of Automation, CAS, China
Cong Cai,  Beijing Institute of Technology, China
Bin Liu,  Institute of Automation, CAS, China
Zheng Lian, Institute of Automation, CAS, China

Contact Us

Please refer to the webpage, http://www.iscslp2024.com/emotionChallenges, for detailed information on the Challenge.
If you have any questions, you can contact us through the official email. IEPR2024@iscslp2024.com


WeChat Group(left) for Challenge Participants：
If you are unable to join the WeChat group, please follow the official WeChat public account of ISCSLP2024(right) and reply “IEPR” to obtain the way to join the group.（Both methods join the same WeChat Group.）