AI Paper Atlas

Research map / 3D vision / 3D generation

MicroDreamer: Efficient 3D Generation in $\sim$20 Seconds by Score-based Iterative Reconstruction

Luxi Chen, Zhengyi Wang, Zihan Zhou, Tingting Gao, Hang Su, Jun Zhu, Chongxuan Li

TPAMI 2025 · 3D vision › 3D generation · approach: Score distillation for text-to-3D · 9 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use SIR instead of SDSNeRF generation · NeRF with MVDream, Stable Zero123, ImageDreamspeedup: 5-20 timesImproved
use MicroDreamer-20 iter instead of DreamGaussian3DGS mesh · Image-to-3D, 87 imagesCLIP similarity: 0.692 -> 0.697Improved
use MicroDreamer-20 iter instead of DreamGaussian3DGS mesh · Image-to-3D, 87 imagesgeneration time: ~30+10s -> ~18+3sImproved
use MicroDreamer-30 iter instead of DreamGaussian3DGS mesh · Image-to-3D, 87 imagesCLIP similarity: 0.692 -> 0.711Improved
use MicroDreamer-20 iter instead of DreamGaussian-300 iter at similar time3DGS mesh · Image-to-3D, 87 imagesCLIP similarity: 0.641 -> 0.697Improved
increase MicroDreamer iterations from 20 to 30MicroDreamer · Image-to-3D, 87 imagesCLIP similarity: 0.697 -> 0.711Improved
optimize SIR in pixel space instead of latent space (ablation)SIR on NeRF · NeRF generationspeedup: 2-3 timesImproved
apply SDS in pixel space (ablation)SDS · 3D generationqualityHurt
use iterative reconstruction (K>0) instead of K=0 initialization only (ablation)SIR · 3DGSgeneration qualityImproved
increase K beyond 30 (ablation)SIR · 3DGSgeneration qualityNo clear effect
use more camera poses than 1 (ablation)SIR · 3DGS with Stable Zero123generation qualityImproved
use linear time schedule instead of random or square schedule (ablation)SIR · 3DGSgeneration qualityNo clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy