MVDream: Multi-view Diffusion for 3D Generation

Yichun Shi, Peng Wang, Jianglong Ye, Mai Long, Kejie Li +1 more
2/12/2026

No official code found. Best implementation: bytedance/MVDream (987 stars, MIT, last commit 2023-10-07).

Abstract

We introduce MVDream, a diffusion model that is able to generate consistent multi-view images from a given text prompt. Learning from both 2D and 3D data, a multi-view diffusion model can achieve the generalizability of 2D diffusion models and the consistency of 3D renderings. We demonstrate that such a multi-view diffusion model is implicitly a generalizable 3D prior agnostic to 3D representations. It can be applied to 3D generation via Score Distillation Sampling, significantly enhancing the consistency and stability of existing 2D-lifting methods. It can also learn new concepts from a few 2D examples, akin to DreamBooth, but for 3D generation.

View on arXivView PDF

Code Implementations(1)

bytedance/MVDreamCommunity71%
MIT

Multi-view Diffusion for 3D Generation

98774PythonOct 3, 20233 years agoMIT
research

Similar Papers

SyncDreamer: Generating Multiview-consistent Images from a Single-view Image

91% match1 repo1.0k2026

ViewFusion: Towards Multi-View Consistency via Interpolated Denoising

90% match1 repo542026

Carve3D: Improving Multi-view Reconstruction Consistency for Diffusion Models with RL Finetuning

90% match1 repo382026

EpiDiff: Enhancing Multi-View Synthesis via Localized Epipolar-Constrained Diffusion

90% match1 repo1432026

Multi-view Image Prompted Multi-view Diffusion for Improved 3D Generation

90% match2024

DreamComposer: Controllable 3D Object Generation via Multi-View Conditions

89% match1 repo1352026

Cite this paper

@article{shi2026mvdream,
  title  = {MVDream: Multi-view Diffusion for 3D Generation},
  author = {Yichun Shi and Peng Wang and Jianglong Ye and Mai Long and Kejie Li and X. Yang},
  year   = {2026},
  eprint = {2308.16512},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/pdf/2308.16512},
  journal = {ICLR 2023 2023}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2308.16512.svg)](https://codeofpaper.com/paper/2308.16512)

Discussion