Empowering Vector Graphics with Consistently Arbitrary Viewing and View-dependent Visibility

TL;DR

This paper introduces Dream3DVG, a novel text-to-vector graphics generation method that enables arbitrary viewpoints, progressive detail refinement, and view-dependent occlusion handling through a dual-branch optimization framework.

Contribution

It presents a dual-branch framework combining 3D Gaussian Splatting and vector graphics optimization, enabling consistent, view-dependent, and detail-controlled vector graphic generation from text prompts.

Findings

Outperforms existing methods on 3D sketches and iconographies

Achieves high cross-view consistency and occlusion-aware rendering

Enables progressive detail control during generation

Abstract

This work presents a novel text-to-vector graphics generation approach, Dream3DVG, allowing for arbitrary viewpoint viewing, progressive detail optimization, and view-dependent occlusion awareness. Our approach is a dual-branch optimization framework, consisting of an auxiliary 3D Gaussian Splatting optimization branch and a 3D vector graphics optimization branch. The introduced 3DGS branch can bridge the domain gaps between text prompts and vector graphics with more consistent guidance. Moreover, 3DGS allows for progressive detail control by scheduling classifier-free guidance, facilitating guiding vector graphics with coarse shapes at the initial stages and finer details at later stages. We also improve the view-dependent occlusions by devising a visibility-awareness rendering module. Extensive results on 3D sketches and 3D iconographies, demonstrate the superiority of the method on different abstraction levels of details, cross-view consistency, and occlusion-aware stroke culling.