DeepPanoContext: Panoramic 3D Scene Understanding with Holistic Scene Context Graph and Relation-based Optimization

TL;DR

DeepPanoContext introduces a novel panoramic 3D scene understanding approach that leverages a holistic scene context graph and relation-based optimization to accurately recover room layouts and object details from a single panorama.

Contribution

The paper proposes a new graph neural network-based context model and a differentiable optimization module for panoramic 3D scene understanding, along with a synthetic dataset for training.

Findings

Outperforms existing methods in geometry accuracy

Achieves better object arrangement results

Demonstrates effectiveness on a new synthetic dataset

Abstract

Panorama images have a much larger field-of-view thus naturally encode enriched scene context information compared to standard perspective images, which however is not well exploited in the previous scene understanding methods. In this paper, we propose a novel method for panoramic 3D scene understanding which recovers the 3D room layout and the shape, pose, position, and semantic category for each object from a single full-view panorama image. In order to fully utilize the rich context information, we design a novel graph neural network based context model to predict the relationship among objects and room layout, and a differentiable relationship-based optimization module to optimize object arrangement with well-designed objective functions on-the-fly. Realizing the existing data are either with incomplete ground truth or overly-simplified scene, we present a new synthetic dataset with good diversity in room layout and furniture placement, and realistic image quality for total panoramic 3D scene understanding. Experiments demonstrate that our method outperforms existing methods on panoramic scene understanding in terms of both geometry accuracy and object arrangement. Code is available at https://chengzhag.github.io/publication/dpc.