DreamReward: Text-to-3D Generation with Human Preference

arXiv:2403.14613v1 Announce Type: cross
Abstract: 3D content creation from text prompts has shown remarkable success recently. However, current text-to-3D methods often generate 3D results that do not align well with human preferences. In this paper, we present a comprehensive framework, coined DreamReward, to learn and improve text-to-3D models from human preference feedback. To begin with, we collect 25k expert comparisons based on a systematic annotation pipeline including rating and ranking. Then, we build Reward3D -- the first general-purpose text-to-3D human …

3d models abstract arxiv cs.cl cs.cv cs.lg current feedback framework generate however human learn paper prompts results success text type