Free-form 3D Scene Inpainting with Dual-stream GAN

TL;DR

This paper introduces a new free-form 3D scene inpainting task and dataset, along with a dual-stream GAN model that effectively handles large missing regions by fusing geometry and color information.

Contribution

The paper presents a novel free-form 3D scene inpainting task, a new dataset FF-Matterport with diverse missing regions, and a dual-stream GAN approach tailored for this challenging problem.

Findings

The proposed method outperforms existing scene completion techniques.

The dual-stream GAN effectively preserves semantic boundaries and details.

Experiments demonstrate the dataset's diversity and the model's robustness.

Abstract

Nowadays, the need for user editing in a 3D scene has rapidly increased due to the development of AR and VR technology. However, the existing 3D scene completion task (and datasets) cannot suit the need because the missing regions in scenes are generated by the sensor limitation or object occlusion. Thus, we present a novel task named free-form 3D scene inpainting. Unlike scenes in previous 3D completion datasets preserving most of the main structures and hints of detailed shapes around missing regions, the proposed inpainting dataset, FF-Matterport, contains large and diverse missing regions formed by our free-form 3D mask generation algorithm that can mimic human drawing trajectories in 3D space. Moreover, prior 3D completion methods cannot perform well on this challenging yet practical task, simply interpolating nearby geometry and color context. Thus, a tailored dual-stream GAN method is proposed. First, our dual-stream generator, fusing both geometry and color information, produces distinct semantic boundaries and solves the interpolation issue. To further enhance the details, our lightweight dual-stream discriminator regularizes the geometry and color edges of the predicted scenes to be realistic and sharp. We conducted experiments with the proposed FF-Matterport dataset. Qualitative and quantitative results validate the superiority of our approach over existing scene completion methods and the efficacy of all proposed components.