Predicting the Attractiveness of Real-Estate Images by Pairwise Comparison using Deep Learning

This paper presents a new user experience for online apartment search using functionality and comfort as query items. Specifically, it has three technical contributions. First, we present a new dataset on the perceived functionality and comfort scores of residential floor plans using nine question statements about the level of comfort, openness, privacy, etc. Second, we propose an algorithm to predict the scores from the floor plan images. Lastly, we implement a new apartment search system and conduct a large-scale usability study using crowdsourcing. The experimental results show that our apartment search system can provide a better user experience. To the best of our knowledge, this is the first work to propose a highly accurate machine learning model for predicting the subjective functionality and comfort of apartments.

show abstract

“…Wang et al [9] predicted which among two images of the same property is more attractive using a pairwise comparison network.…”

Section: A Real Estate Tasks Using Property Imagesmentioning

confidence: 99%

Subjective Functionality and Comfort Prediction for Apartment Floor Plans and Its Application to Intuitive Online Property Search

Narahara

Yamasaki

2023

IEEE Trans. Multimedia

View full text Add to dashboard Cite

show abstract

“…Several previous works have studied subjective evaluations in architecture using 2-D images, such as property images, photos, satellite images, and floor plan images (Wang, et al, 2019, Narahara andYamasaki, 2019). In terms of urban city-scale applications, there are some standardized quantified measures for urban living conditions, including some qualitative, subjective criteria such as safety and ease of access (Walk Score, Zillow Inc. 2021).…”

Section: Introductionmentioning

confidence: 99%

Kurashiki Viewer: Qualitative Evaluations of Architectural Spaces inside Virtual Reality

Narahara¹

2022

CAADRIA Proceedings

View full text Add to dashboard Cite

This paper discusses how virtual reality (VR) environments can be employed as a data collection tool beyond visualization and representation tools through a simple experiment in a VR space and speculates about its potential applications. Using a VR model that runs on a web browser based on an existing historic town in Japan called Kurashiki, the experiment asked 30 recruited participants to freely walk around and leave ratings on a 5-point scale on any buildings or objects appealing to them. The proposed system in this paper can display points of interest of multiple participants using heatmaps superimposed on a map that can help users visually understand statistical preferences among them. The project's goal is to provide a quantitative means for qualitative values of architectural and urban spaces, making such data more shareable. We intended to show that such a platform could help multiple stakeholders reach better consensuses and possibly collect training datasets for machine learning models that could extract features related to the attractiveness in architecture and urban spaces.

show abstract

A Deep Learning-Based Segregation of Housing Image Data for Real Estate Application

Kumari

Maan

Sangwan

2021

Intelligent Learning for Computer Vision

View full text Add to dashboard Cite

Predicting the Attractiveness of Real-Estate Images by Pairwise Comparison using Deep Learning

Cited by 5 publications

References 20 publications

Subjective Functionality and Comfort Prediction for Apartment Floor Plans and Its Application to Intuitive Online Property Search

Subjective Functionality and Comfort Prediction for Apartment Floor Plans and Its Application to Intuitive Online Property Search

Kurashiki Viewer: Qualitative Evaluations of Architectural Spaces inside Virtual Reality

A Deep Learning-Based Segregation of Housing Image Data for Real Estate Application

Contact Info

Product

Resources

About