🌍 Model Card for City Identification using DL-SLICER-models

This model card describes the DL-SLICER deep learning tool, which is designed for satellite-based city identification and feature analysis.

Figure 1: Site images from satellite data, representing cities by their International Air Transport Association (IATA) codes, used for city identification model training.

It has been generated using DL-SLICER dataset.

Model Details

Model Description

The models are based on the ResNet architecture and are trained on satellite imagery to classify and distinguish between 45 cities worldwide. The tool utilizes an Explainable AI method, specifically Relevance Class Activation Maps (CAMs), to identify the visual features that characterize each city. The DL-SLICER repository contains four ResNet models (resnet-18, resnet-34, resnet-50, and resnet-101), each with two sets of checkpoints: best.pth (representing the best-performing model) and last.pth (the final model checkpoint).

• Developed by: Ulzhan Bissarinova, Aidana Tleuken, Sofiya Alimukhambetova, Huseyin Atakan Varol, Ferhat Karaca

Model Sources

• Repository: https://github.com/IS2AI/city-identification
• Paper: Bissarinova, Ulzhan, Aidana Tleuken, Sofiya Alimukhambetova, Huseyin Atakan Varol, and Ferhat Karaca. 2024. "DL-SLICER: Deep Learning for Satellite-Based Identification of Cities with Enhanced Resemblance" Buildings 14, no. 2: 551. https://doi.org/10.3390/buildings14020551

Uses

The DL-SLICER models are intended for researchers, policymakers, city managers, and urban planners. The models can be used to:

• Identify similar cities based on satellite data patterns.

• Analyze salient urban features for urban planning, crisis management, and economic policy decisions.

• Provide data for indices and concepts related to sustainability and smart cities.

Citation

If you use the dataset/source code/pre-trained models in your research, please cite our work

BibTeX:

@Article{buildings14020551,
AUTHOR = {Bissarinova, Ulzhan and Tleuken, Aidana and Alimukhambetova, Sofiya and Varol, Huseyin Atakan and Karaca, Ferhat},
TITLE = {DL-SLICER: Deep Learning for Satellite-Based Identification of Cities with Enhanced Resemblance},
JOURNAL = {Buildings},
VOLUME = {14},
YEAR = {2024},
NUMBER = {2},
ARTICLE-NUMBER = {551},
URL = {https://www.mdpi.com/2075-5309/14/2/551},
ISSN = {2075-5309},
ABSTRACT = {This paper introduces a deep learning (DL) tool capable of classifying cities and revealing the features that characterize each city from a visual perspective. The study utilizes city view data captured from satellites and employs a methodology involving DL-based classification for city identification, along with an Explainable Artificial Intelligence (AI) tool to unveil definitive features of each city considered in this study. The city identification model implemented using the ResNet architecture yielded an overall accuracy of 84%, featuring 45 cities worldwide with varied geographic locations, Human Development Index (HDI), and population sizes. The portraying attributes of urban locations have been investigated using an explanatory visualization tool named Relevance Class Activation Maps (CAM). The methodology and findings presented by the current study enable decision makers, city managers, and policymakers to identify similar cities through satellite data, understand the salient features of the cities, and make decisions based on similarity patterns that can lead to effective solutions in a wide range of objectives such as urban planning, crisis management, and economic policies. Analyzing city similarities is crucial for urban development, transportation strategies, zoning, improvement of living conditions, fostering economic success, shaping social justice policies, and providing data for indices and concepts such as sustainability and smart cities for urban zones sharing similar patterns.},
DOI = {10.3390/buildings14020551}
}