Sketch-A-Shape: Zero-Shot Sketch-to-3D Shape Generation

TL;DR

This paper introduces Sketch-A-Shape, a zero-shot method that leverages pre-trained vision models to generate 3D shapes from sketches without needing paired datasets, effectively handling various sketch abstractions.

Contribution

It demonstrates that features from large pre-trained vision models can be used to generate 3D shapes from sketches in a zero-shot manner, bypassing the need for paired training data.

Findings

Effective zero-shot 3D shape generation from sketches.

Robustness to different levels of sketch abstraction.

No requirement for paired sketch-shape datasets during training.

Abstract

Significant progress has recently been made in creative applications of large pre-trained models for downstream tasks in 3D vision, such as text-to-shape generation. This motivates our investigation of how these pre-trained models can be used effectively to generate 3D shapes from sketches, which has largely remained an open challenge due to the limited sketch-shape paired datasets and the varying level of abstraction in the sketches. We discover that conditioning a 3D generative model on the features (obtained from a frozen large pre-trained vision model) of synthetic renderings during training enables us to effectively generate 3D shapes from sketches at inference time. This suggests that the large pre-trained vision model features carry semantic signals that are resilient to domain shifts, i.e., allowing us to use only RGB renderings, but generalizing to sketches at inference time. We conduct a comprehensive set of experiments investigating different design factors and demonstrate the effectiveness of our straightforward approach for generation of multiple 3D shapes per each input sketch regardless of their level of abstraction without requiring any paired datasets during training.