Datasetpublic

Rayford295/BiTemporal-StreetView-Damage

Each folder contains bi-temporal street-view image pairs representing the same or nearby locations before and after a disaster event. 🏷️ Damage Severity Labels Folder Severity Level Description folder_0 Mild Minor visible damage (e.g., small debris, fallen branches, limited facade damage) folder_1 Moderate Clearly visible structural or environmental damage folder_2 Severe Extensive or catastrophic damage, including near-complete destruction These labels are… See the full description on the dataset page: https://huggingface.co/datasets/Rayford295/BiTemporal-StreetView-Damage.

sourceHugging Faceupdated 8mo agoView on Hugging Face

0likes41downloads

Dataset Card

Each folder contains bi-temporal street-view image pairs representing the same or nearby locations before and after a disaster event.

🏷️ Damage Severity Labels

Folder	Severity Level	Description
`folder_0`	Mild	Minor visible damage (e.g., small debris, fallen branches, limited facade damage)
`folder_1`	Moderate	Clearly visible structural or environmental damage
`folder_2`	Severe	Extensive or catastrophic damage, including near-complete destruction

These labels are designed to capture perceived damage severity from a street-level viewpoint, aligning with how humans visually assess disaster impacts.

🔍 Data Source & Annotation

—Street-view images were collected from publicly available street-level imagery platforms.
—Each bi-temporal image pair was manually reviewed.
—Damage severity labels were assigned based on post-disaster visual evidence, with reference to the pre-disaster condition when available.
—The dataset prioritizes relative change and perceptual severity, rather than pixel-level structural measurements.

🎯 Intended Use

This dataset can be used for:

—Bi-temporal damage severity classification
—Disaster perception modeling
—Vision-only and vision-language disaster assessment
—Benchmarking deep learning and foundation models
—GeoAI research on post-disaster urban damage

⚠️ Limitations

—Labels are provided at the image-pair level, not at the pixel level.
—Viewpoint differences, occlusions, and lighting variations may exist between image pairs.
—The dataset focuses on hurricane-related damage and may not directly generalize to other disaster types.

📜 License

This dataset is released under the Creative Commons Attribution-NonCommercial 4.0 (CC BY-NC 4.0) license.

—✔ Free for academic and non-commercial research
—❌ Commercial use is not permitted

Proper citation is required for any use of this dataset.

📖 Citation

If you use this dataset in your research, please cite:

bibtex

@dataset{yang_bitemporal_streetview_damage_2026,
  author = {Yang, Yifan},
  title = {BiTemporal-StreetView-Damage: A Bi-Temporal Street-View Dataset for Post-Disaster Damage Severity Classification},
  year = {2026},
  publisher = {Hugging Face},
  url = {https://huggingface.co/datasets/Rayford295/BiTemporal-StreetView-Damage}
}

and

@article{YANG2025102335,
title = {Hyperlocal disaster damage assessment using bi-temporal street-view imagery and pre-trained vision models},
journal = {Computers, Environment and Urban Systems},
volume = {121},
pages = {102335},
year = {2025},
issn = {0198-9715},
doi = {https://doi.org/10.1016/j.compenvurbsys.2025.102335},
url = {https://www.sciencedirect.com/science/article/pii/S0198971525000882},
author = {Yifan Yang and Lei Zou and Bing Zhou and Daoyang Li and Binbin Lin and Joynal Abedin and Mingzheng Yang},
keywords = {Disaster resilience, Street-view imagery, Dual-channel neural network, Pre-trained vision model, Damage estimation}
}