Are We Ready for RL in Text-to-3D Generation? A Progressive Investigation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Tang, Yiwen, Guo, Zoey, Zhu, Kaixin, Zhang, Ray, Chen, Qizhi, Jiang, Dongzhi, Liu, Junli, Zeng, Bohan, Song, Haoming, Qu, Delin, Bai, Tianyi, Xu, Dan, Zhang, Wentao, Zhao, Bin
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915669936701440
author Tang, Yiwen
Guo, Zoey
Zhu, Kaixin
Zhang, Ray
Chen, Qizhi
Jiang, Dongzhi
Liu, Junli
Zeng, Bohan
Song, Haoming
Qu, Delin
Bai, Tianyi
Xu, Dan
Zhang, Wentao
Zhao, Bin
author_facet Tang, Yiwen
Guo, Zoey
Zhu, Kaixin
Zhang, Ray
Chen, Qizhi
Jiang, Dongzhi
Liu, Junli
Zeng, Bohan
Song, Haoming
Qu, Delin
Bai, Tianyi
Xu, Dan
Zhang, Wentao
Zhao, Bin
contents Reinforcement learning (RL), earlier proven to be effective in large language and multi-modal models, has been successfully extended to enhance 2D image generation recently. However, applying RL to 3D generation remains largely unexplored due to the higher spatial complexity of 3D objects, which require globally consistent geometry and fine-grained local textures. This makes 3D generation significantly sensitive to reward designs and RL algorithms. To address these challenges, we conduct the first systematic study of RL for text-to-3D autoregressive generation across several dimensions. (1) Reward designs: We evaluate reward dimensions and model choices, showing that alignment with human preference is crucial, and that general multi-modal models provide robust signal for 3D attributes. (2) RL algorithms: We study GRPO variants, highlighting the effectiveness of token-level optimization, and further investigate the scaling of training data and iterations. (3) Text-to-3D Benchmarks: Since existing benchmarks fail to measure implicit reasoning abilities in 3D generation models, we introduce MME-3DR. (4) Advanced RL paradigms: Motivated by the natural hierarchy of 3D generation, we propose Hi-GRPO, which optimizes the global-to-local hierarchical 3D generation through dedicated reward ensembles. Based on these insights, we develop AR3D-R1, the first RL-enhanced text-to-3D model, expert from coarse shape to texture refinement. We hope this study provides insights into RL-driven reasoning for 3D generation. Code is released at https://github.com/Ivan-Tang-3D/3DGen-R1.
format Preprint
id arxiv_https___arxiv_org_abs_2512_10949
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Are We Ready for RL in Text-to-3D Generation? A Progressive Investigation
Tang, Yiwen
Guo, Zoey
Zhu, Kaixin
Zhang, Ray
Chen, Qizhi
Jiang, Dongzhi
Liu, Junli
Zeng, Bohan
Song, Haoming
Qu, Delin
Bai, Tianyi
Xu, Dan
Zhang, Wentao
Zhao, Bin
Computer Vision and Pattern Recognition
Artificial Intelligence
Computation and Language
Reinforcement learning (RL), earlier proven to be effective in large language and multi-modal models, has been successfully extended to enhance 2D image generation recently. However, applying RL to 3D generation remains largely unexplored due to the higher spatial complexity of 3D objects, which require globally consistent geometry and fine-grained local textures. This makes 3D generation significantly sensitive to reward designs and RL algorithms. To address these challenges, we conduct the first systematic study of RL for text-to-3D autoregressive generation across several dimensions. (1) Reward designs: We evaluate reward dimensions and model choices, showing that alignment with human preference is crucial, and that general multi-modal models provide robust signal for 3D attributes. (2) RL algorithms: We study GRPO variants, highlighting the effectiveness of token-level optimization, and further investigate the scaling of training data and iterations. (3) Text-to-3D Benchmarks: Since existing benchmarks fail to measure implicit reasoning abilities in 3D generation models, we introduce MME-3DR. (4) Advanced RL paradigms: Motivated by the natural hierarchy of 3D generation, we propose Hi-GRPO, which optimizes the global-to-local hierarchical 3D generation through dedicated reward ensembles. Based on these insights, we develop AR3D-R1, the first RL-enhanced text-to-3D model, expert from coarse shape to texture refinement. We hope this study provides insights into RL-driven reasoning for 3D generation. Code is released at https://github.com/Ivan-Tang-3D/3DGen-R1.
title Are We Ready for RL in Text-to-3D Generation? A Progressive Investigation
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2512.10949