9th Emotion Recognition in the Wild Challenge (EmotiW)
ACM International Conference on Multimodal Interaction 2023, Paris
ACM International Conference on Multimodal Interaction 2023, Paris
a. EngageNet based Engagement Prediction - The task is to predict engagement level of participant from an observer's perspective. This is a four class classification task. The details regarding the dataset can be accessed at [LINK]
Baseline - Pre-trained model Masked Autoencoder for facial video Representation LearnINg (MARLIN) model [LINK] has been used to extract features. A MLP was trained with MARLIN as input and achieved 65.54% on the Validation set.
b. The audio-video group emotion recognition challenge contains group videos downloaded from YouTube with creative commons license. The data has lot of variations in terms of context, number of people, video quality, etc. The task is to classify each video in three classes - positive, neutral and negative.