Novel CNN-Based Approach for Reading Urban Form Data in 2D Images: An Application for Predicting Restaurant Location in Seoul, Korea

被引:6
|
作者
Yang, Jeyun [1 ]
Kwon, Youngsang [1 ,2 ]
机构
[1] Seoul Natl Univ, Dept Civil & Environm Engn, Seoul 08826, South Korea
[2] Seoul Natl Univ, Adv Inst Convergence Technol, Smart City Res Ctr, Suwon 16229, South Korea
基金
新加坡国家研究基金会;
关键词
convolutional neural network; CNN; restaurant location; retail location; urban retail store; deep learning; classification model; RETAIL LOCATION; CITIES; GENTRIFICATION; MORPHOLOGY; FRAMEWORK; DEMAND; FOOD;
D O I
10.3390/ijgi12090373
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Artificial intelligence (AI) has demonstrated its ability to complete complex tasks in various fields. In urban studies, AI technology has been utilized in some limited domains, such as control of traffic and air quality. This study uses AI to better understand diverse urban studies data through a novel approach that uses a convolutional neural network (CNN). In this study, a building outline in the form of a two-dimensional image is used with its corresponding metadata to test the applicability of CNN in reading urban data. MobileNet, a high-efficiency CNN model, is trained to predict the location of restaurants in each building in Seoul, Korea. Consequently, using only 2D image data, the model satisfactorily predicts the locations of restaurants (AUC = 0.732); the model with 2D images and their metadata has higher performance but has an overfitting problem. In addition, the model using only 2D image data accurately predicts the regional distribution of restaurants and shows some typical urban forms with restaurants. The proposed model has several technical limitations but shows the potential to provide a further understanding of urban settings.
引用
收藏
页数:21
相关论文
共 6 条
  • [1] Extending 2D Saliency Models for Head Movement Prediction in 360-degree Images using CNN-based Fusion
    Djemai, Ibrahim
    Fezza, Sid Ahmed
    Hamidouche, Wassim
    Deforges, Olivier
    2020 IEEE INTERNATIONAL SYMPOSIUM ON CIRCUITS AND SYSTEMS (ISCAS), 2020,
  • [2] A novel 3D lumbar vertebrae location and segmentation method based on the fusion envelope of 2D hybrid visual projection images
    Wu, Zhengyang
    Xia, Guifeng
    Zhang, Xiaoheng
    Zhou, Fayuan
    Ling, Jing
    Ni, Xin
    Li, Yongming
    COMPUTERS IN BIOLOGY AND MEDICINE, 2022, 151
  • [3] DEEP-COVID-SEV: AN ENSEMBLE 2D AND 3D CNN-BASED APPROACH FOR COVID-19 SEVERITY PREDICTION FROM 3D CT-SCANS
    Bougourzi, Fares
    Dornaika, Fadi
    Nakib, Amir
    Distante, Cosimo
    Taleb-Ahmed, Abdelmalik
    2023 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING WORKSHOPS, ICASSPW, 2023,
  • [4] Deep CNN-based classification of motor imagery tasks from EEG signals using 2D wavelet transformed images of adaptively reconstructed signals from MVMD decomposed modes
    Alizadeh, N.
    Afrakhteh, S.
    Mosavi, M. R.
    INTERNATIONAL JOURNAL OF IMAGING SYSTEMS AND TECHNOLOGY, 2023, 33 (06) : 1988 - 2011
  • [5] Hybrid-COVID: a novel hybrid 2D/3D CNN based on cross-domain adaptation approach for COVID-19 screening from chest X-ray images
    Khaled Bayoudh
    Fayçal Hamdaoui
    Abdellatif Mtibaa
    Physical and Engineering Sciences in Medicine, 2020, 43 : 1415 - 1431
  • [6] Hybrid-COVID: a novel hybrid 2D/3D CNN based on cross-domain adaptation approach for COVID-19 screening from chest X-ray images
    Bayoudh, Khaled
    Hamdaoui, Faycal
    Mtibaa, Abdellatif
    PHYSICAL AND ENGINEERING SCIENCES IN MEDICINE, 2020, 43 (04) : 1415 - 1431