VD-GR: Boosting Visual Dialog with Cascaded Spatial-Temporal Multi-Modal GRaphs.

Adnen Abdessaied Lei Shi Andreas Bulling

Published in: CoRR (2023)

Keyphrases

multi modal
spatial temporal
cross modal
video search
spatio temporal
spatial and temporal
action recognition
temporal information
multi modality
single modality
video shots
image annotation
high dimensional
visual information
visual features
uni modal
semantic concepts
multiple modalities
co occurrence