Layouts from Panoramic Images with Geometry and Deep Learning

TL;DR

This paper introduces a method for 3D indoor scene layout recovery from single panoramic images by integrating geometric reasoning with deep learning, effectively handling both simple and complex room shapes.

Contribution

It presents a novel approach combining geometry and deep learning for accurate 3D layout estimation from panoramic images, outperforming existing methods.

Findings

Effective layout recovery for simple and complex rooms

Improved accuracy over state-of-the-art methods

Validated on SUN360 and Stanford datasets

Abstract

In this paper, we propose a novel procedure for 3D layout recovery of indoor scenes from single 360 degrees panoramic images. With such images, all scene is seen at once, allowing to recover closed geometries. Our method combines strategically the accuracy provided by geometric reasoning (lines and vanishing points) with the higher level of data abstraction and pattern recognition achieved by deep learning techniques (edge and normal maps). Thus, we extract structural corners from which we generate layout hypotheses of the room assuming Manhattan world. The best layout model is selected, achieving good performance on both simple rooms (box-type) and complex shaped rooms (with more than four walls). Experiments of the proposed approach are conducted within two public datasets, SUN360 and Stanford (2D-3D-S) demonstrating the advantages of estimating layouts by combining geometry and deep learning and the effectiveness of our proposal with respect to the state of the art.