Unveiling the Power of Audio-Visual Early Fusion Transformers with Dense Interactions through Masked Modeling

Open in new window