Contrastive preference optimization, combining ImageReward and PickScore with static and LLM-generated negative prompts, improves semantic alignment in SDS-based 3D human generation, especially for long prompts.
TADA! Text to Animatable Digital Avatars,
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
-
Text-driven 3D Human Generation via Contrastive Preference Optimization
Contrastive preference optimization, combining ImageReward and PickScore with static and LLM-generated negative prompts, improves semantic alignment in SDS-based 3D human generation, especially for long prompts.