Weakly supervised fitting recovers consistent pseudo-skeletons from mesh sequences, then a text-conditioned transformer with Motion-GRPO generates editable skeleton-driven 4D mesh animations.
Learning transferable visual models from natural language supervision
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CV 2years
2026 2representative citing papers
DDiffusion uses semantic retrieval on prompt embeddings and localized editing inside the diffusion process to suppress NSFW content while avoiding binary allow/block signals.
citing papers explorer
-
SkelGen4D: Weakly-Supervised Skeleton-Based 4D Generation for Text-Driven Mesh Animation
Weakly supervised fitting recovers consistent pseudo-skeletons from mesh sequences, then a text-conditioned transformer with Motion-GRPO generates editable skeleton-driven 4D mesh animations.
-
Disciplined Diffusion: Text-to-Image Diffusion Model against NSFW Generation
DDiffusion uses semantic retrieval on prompt embeddings and localized editing inside the diffusion process to suppress NSFW content while avoiding binary allow/block signals.