REACT 2026 Challenge: The Fourth Personalised Multiple Appropriate Facial Reaction Generation in Dyadic Interactions

Siyang Song, Micol Spitale, Zijian Wu, Xiangyu Kong, Cheng Luo, Cristina Palmero, German Barquero, Sergio Escalera, Michel Valstar, Mohamed Daoudi, Fabien Ringeval, Andrew Howes, Elisabeth Andre, Hatice Gunes

Keywords: Multiple Appropriate Facial Reaction Generation, Human Behaviour Understanding, Affective Computing, Generative AI

Abstract: According to the Stimulus Organism Response (SOR) theory, for a given external stimulus, individuals may react differently according to their internal state and external contextual factors in a specific period in time. Analogously, in dyadic interactions, a broad spectrum of human facial reactions might be appropriate for responding to a specific human speaker behaviour. Following the successful organisation of the REACT 2023, REACT 2024 and REACT 2025 challenge series, a body of generative deep learning (DL) models have been investigated for the problem of multiple appropriate facial reaction generation (MAFRG). While REACT 2023 and 2024 challenges were built on human-human dyadic interaction datasets collected for other purposes, the REACT 2025 challenge provided the first natural and large-scale audio-visual Multiple Appropriate Facial Reaction Generation (MAFRG) dataset (called MARS) recording 137 human-human dyadic interactions containing a total of 3,105 interaction sessions covering five different topics. This year, we are proposing the REACT 2026 challenge encouraging the development and benchmarking of Machine Learning (ML) models that can be used to generate multiple appropriate, diverse, realistic and synchronised human-style facial reactions expressed by human listeners in response to each input speaker behaviour expressed by the corresponding speaker. As a key of the challenge, we will continuously provide challenge participants with MARS dataset but additionally providing individual-level Big-Five personality labels and EEG recordings. This introduces a new one-to-many personalised reaction generation setting combining behavioural, affective and neurophysiological signals, which remains largely unexplored in current dyadic interaction modelling. We will then invite the challenge participating groups to submit their developed / trained ML models for evaluation, which will be benchmarked in terms of the appropriateness, diversity, realism and synchronisation of their generated facial reactions.

Email Sharing: We authorize the sharing of all author emails with Program Chairs.

Data Release: We authorize the release of our submission and author names to the public in the event of acceptance.

Submission Number: 18