REVIVE 3D: Refinement via Encoded Voluminous Inflated prior for Volume Enhancement

arXiv cs.CV / 5/1/2026

📰 NewsIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper proposes REVIVE 3D, a two-stage, plug-and-play generative pipeline that produces voluminous 3D assets from flat 2D images, addressing the lack of 3D cues in such inputs.
Stage 1 builds an “Inflated Prior” by inflating the foreground silhouette to recover global volume while adding part-aware details to preserve local structure.
Stage 2 introduces “3D Latent Refinement,” which injects Gaussian noise into the prior’s latent representation and then denoises it using geometric cues to tap into the backbone’s pretrained 3D knowledge.
The framework also supports image-conditioned 3D editing, and the authors introduce Compactness and Normal Anisotropy metrics that correlate with human perception of volume and surface quality.
Experiments on a challenging flat-image dataset show state-of-the-art results with both qualitative and quantitative evaluations, validated further via a user study.

Abstract

Recent generative models have shown strong performance in generating diverse 3D assets from 2D images, a fundamental research topic in computer vision and graphics. However, these models still struggle to generate voluminous 3D assets when the input is a flat image that provides limited 3D cues. We introduce REVIVE 3D, a two-stage, plug-and-play pipeline for generating voluminous 3D assets from flat images. In Stage 1, we construct an Inflated Prior by inflating the foreground silhouette to recover global volume and superimposing part-aware details to capture local structure. In Stage 2, 3D Latent Refinement injects Gaussian noise into the Inflated Prior's latent and then denoises it, using the prior's geometric cues to leverage the backbone's pretrained 3D knowledge. Furthermore, our framework supports image-conditioned 3D editing. To quantify volume and surface flatness, we propose Compactness and Normal Anisotropy. We validate Compactness and Normal Anisotropy through a user study, showing that these metrics align with human perception of volume and quality. We show that REVIVE 3D achieves state-of-the-art performance on a challenging flat image dataset, based on extensive qualitative and quantitative evaluations.