Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Instant3D: Fast Text-to-3D with Sparse-View Generation and Large Reconstruction Model

About

Text-to-3D with diffusion models has achieved remarkable progress in recent years. However, existing methods either rely on score distillation-based optimization which suffer from slow inference, low diversity and Janus problems, or are feed-forward methods that generate low-quality results due to the scarcity of 3D training data. In this paper, we propose Instant3D, a novel method that generates high-quality and diverse 3D assets from text prompts in a feed-forward manner. We adopt a two-stage paradigm, which first generates a sparse set of four structured and consistent views from text in one shot with a fine-tuned 2D text-to-image diffusion model, and then directly regresses the NeRF from the generated images with a novel transformer-based sparse-view reconstructor. Through extensive experiments, we demonstrate that our method can generate diverse 3D assets of high visual quality within 20 seconds, which is two orders of magnitude faster than previous optimization-based methods that can take 1 to 10 hours. Our project webpage: https://jiahao.ai/instant3d/.

Jiahao Li, Hao Tan, Kai Zhang, Zexiang Xu, Fujun Luan, Yinghao Xu, Yicong Hong, Kalyan Sunkavalli, Greg Shakhnarovich, Sai Bi• 2023

Related benchmarks

TaskDatasetResultRank
View SynthesisGSO (test)
PSNR26.54
19
View SynthesisABO (test)
PSNR27.5
18
3D Material Refinement PreferenceObjaverse
GPT Evaluation Score38.5
8
Multiview text-to-image inpaintingMultiview Inpainting Benchmark 500 images (test)
CLIP Similarity (L)28.78
7
Multi-View ReconstructionDreamFusion (test)
Avg MRC0.0892
7
Text-to-3D GenerationText-to-3D prompts
CLIP Similarity (ViT-L/14)26.87
4
Showing 6 of 6 rows

Other info

Follow for update