Aff-Wild-2

The 'Aff-Wild-2' dataset is a benchmark that contains annotated videos for evaluating valence-arousal estimation in human emotions within naturalistic environments.

Papers using Aff-Wild-2 (3)

Distance-aware Soft Prompt Learning for Multimodal Valence-Arousal Estimation2026Multimodal Alignment With Cross-attentive Grus For Fine-grained Video Understanding2025Multimodal Fusion with Pre-Trained Model Features in Affective Behaviour Analysis In-the-wild2024