Aesthetic quality assessment for real estate images through deep learning methods

Download
2022-12-12
Uçan, Nazlı Özge
In this thesis, we aim to find the aesthetic quality of real estate images. Although aesthetic assessment is a subjective terminology, it is highly correlated with photographic rules. The aesthetic quality of images in real estate affects the decision of potential people of interest. The aesthetic evaluation of images is established via the Aesthetic Visual Assessment (AVA) dataset benchmark. Although AVA is a publicly available and diverse image dataset, it cannot be adapted to the real estate domain. Therefore, we constructed the Real-Estate Aesthetics Assessment Dataset (RAAD), which consists of real and synthetic real estate images. In order to gather subjective user data on the RAAD, a user study is conducted on a custom web-based scoring platform, serving RAAD image data. We analyzed several different methods involving classical vision classifiers and deep image classification models in order to assign an aesthetic quality score to the given real estate image. The results of those different approaches are presented comparatively on the RAAD data.

Suggestions

Improving Visual Design Perception by an Integrated Mixed Reality Environment for Performative Architecture
Akın, Şahin; Ergün, Oğuzcan; Gürsel Dino, Ipek; Sürer, Elif (null; 2019-05-03)
Expressive design visualization is critical for the architectural designers`designers`visual perception of 3D architectural models. Virtual environments, especially mixed reality (MR), have the potential to enhance designers' design perception and spatial understanding. In performance-based design, building simulations and quantitative analysis of performance predictions are integral to the design process. Specifically, daylighting simulation is considered as a crucial operation through the design developme...
Cost Efficiency Analyses of Optimally Sized Steel Frames for Economical Design of Medium-Rise Buildings
Hasançebi, Oğuzhan (2014-09-01)
The study, in this paper, investigates cost-efficient steel frameworks for economical design of medium-rise buildings. A total of 13 steel frames that incorporate various types of beam-column connection and bracing configuration are considered for detailed and comparative cost analyses. The numerical applications are performed using a 10-story building stiffened according to each of these steel frameworks to yield 13 test frames. First design optimizations are carried out in which the total member weight in...
GEOMETRY-CONSTRAINED SPATIAL PYRAMID ADAPTATION FOR IMAGE CLASSIFICATION
Tasli, H. Emrah; Sicre, Ronan; Gevers, Theo; Alatan, Abdullah Aydın (2014-10-30)
This paper proposes a geometry-constrained spatial pyramid adaptation approach for the image classification task. Scene geometry is used as an input parameter for generating the spatial pyramid definitions. The resulting region adaptation is performed in accordance with the predefined geometric guidelines and underlying image characteristics. Using an approximate global geometric correspondence, exploits the idea that images of the same category share a spatial similarity. This assumption is evaluated and j...
Yönetim modalarının yüksek lisans ve doktora tezlerine yansımaları: Bibliyometrik bir analiz
Armutlu, Can; Sağlam Arı, Güler (Orta Doğu Teknik Üniversitesi (Ankara, Turkey), 2010-4)
Çalışmanın amacı akademik bilgi üretiminin bir boyutunu oluşturan tezlerde yönetim modalarına gösterilen ilgiyi tespit etmektir. Bu amaçla Yüksek Öğretim Kurulu internet sayfasında yer alan Ulusal Tez Merkezi üzerinden İşletme bilim dalı altında toplam kalite yönetimi, kalite çemberleri, değişim mühendisliği, kıyaslama, personel güçlendirme ve altı sigma konularında yazılan lisansüstü tezler bibliyometrik analiz yöntemi ile değerlendirilmiştir. Araştırma amacı doğrultusunda belirlenen yönetim modalarını içe...
Analysis of the limits of representation of architectural photographic images in periodicals
Bakht, Nazlı; Güzer, Celal Abdi; Department of Architecture (2007)
This study aims to analyze the limits of representation of architectural photographic images in periodicals within architectural media that shape today’s architectural production. This aim is accomplished in three sections; “examining the power of image in architecture”; “examining architectural media” and “examining the periodicals’ attitudes towards using architectural photography”. A case study is done to uncover the relationship between the photograph and the medium that it takes place on and determine ...
Citation Formats
N. Ö. Uçan, “Aesthetic quality assessment for real estate images through deep learning methods,” M.S. - Master of Science, Middle East Technical University, 2022.