STG-Mamba generates dance videos from music using a spatial-temporal graph Mamba block for skeleton generation and forward-backward self-supervised losses for video synthesis, reporting SOTA on benchmarks.
Audio to body dynamics,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Spatial-Temporal Graph Mamba for Music-Guided Dance Video Synthesis
STG-Mamba generates dance videos from music using a spatial-temporal graph Mamba block for skeleton generation and forward-backward self-supervised losses for video synthesis, reporting SOTA on benchmarks.