Dyadformer: A Multi-modal Transformer for Long-Range Modeling of Dyadic Interactions.
David CurtoAlbert ClapésJavier SelvaSorina SmeureanuJúlio C. S. Jacques JúniorDavid Gallardo-PujolGeorgina GuileraDavid LeivaThomas B. MoeslundSergio EscaleraCristina PalmeroPublished in: ICCVW (2021)
Keyphrases
- multi modal
- long range
- long range interactions
- short range
- multi modality
- conditional random fields
- semantic concepts
- cross modal
- long range correlations
- audio visual
- uni modal
- wireless communication
- machine learning
- neural network
- video search
- high dimensional
- feature space
- multiple modalities
- similarity measure
- image sequences
- rna structures