JigsawPlan: Room Layout Jigsaw Puzzle Extreme Structure from Motion using Diffusion Models

11/24/2022
by   Sepidehsadat Hosseini, et al.
0

This paper presents a novel approach to the Extreme Structure from Motion (E-SfM) problem, which takes a set of room layouts as polygonal curves in the top-down view, and aligns the room layout pieces by estimating their 2D translations and rotations, akin to solving the jigsaw puzzle of room layouts. The biggest discovery and surprise of the paper is that the simple use of a Diffusion Model solves this challenging registration problem as a conditional generation process. The paper presents a new dataset of room layouts and floorplans for 98,780 houses. The qualitative and quantitative evaluations demonstrate that the proposed approach outperforms the competing methods by significant margins.

READ FULL TEXT

page 1

page 3

page 6

page 7

page 8

research
06/01/2022

Semantic Room Wireframe Detection from a Single View

Reconstruction of indoor surfaces with limited texture information or wi...
research
11/23/2022

HouseDiffusion: Vector Floorplan Generation via a Diffusion Model with Discrete and Continuous Denoising

The paper presents a novel approach for vector-floorplan generation via ...
research
08/07/2021

On the optimal layout of a dining room in the era of COVID-19 using mathematical optimization

We consider the problem of maximizing the number of people that a dining...
research
08/19/2019

Floor-SP: Inverse CAD for Floorplans by Sequential Room-wise Shortest Path

This paper proposes a new approach for automated floorplan reconstructio...
research
06/24/2020

Towards Adversarial Planning for Indoor Scenes with Rotation

In this paper, we propose an adversarial model for producing furniture l...
research
06/21/2023

Polygon Detection for Room Layout Estimation using Heterogeneous Graphs and Wireframes

This paper presents a neural network based semantic plane detection meth...
research
11/29/2018

DuLa-Net: A Dual-Projection Network for Estimating Room Layouts from a Single RGB Panorama

We present a deep learning framework, called DuLa-Net, to predict Manhat...

Please sign up or login with your details

Forgot password? Click here to reset