ZAHA Dataset Overview

Dataset Introduction

ZAHA is presently the largest benchmark for point‑cloud façade semantic segmentation.

Highlights

• 601 million labeled points
• Introduces LoFG (Facade Generalisation Level) for hierarchical façade understanding
• Covers multiple architectural styles
• Provides local and global (UTM) coordinate reference systems
• Filenames reference official Bavarian CityGML LoD2 building models
• Includes configuration files for adding custom data

Segmentation Results

LoFG2 Results

LoFG3 Results

Download

• Download Link
• Password: zahahadid

Citation

@article{wysockietalZAHA,
  author = {Wysocki, O. and Tan, Y. and Froech, T. and Xia, Y. and Wysocki, M. and Hoegner, L. and Cremers, D. and Holst, C.},
  title = {ZAHA: Introducing the Level of Facade Generalization and the Large‑Scale Point Cloud Facade Semantic Segmentation Benchmark Dataset},
  booktitle = {Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)},
  year = {2025},
}