ZuriPano

<h1 align="center" style="font-size: 2.8em; margin: 0.4em 0;">🗃️ ZüriPano Dataset</h1> <p align="center"> <a title="Github" href="https://github.com/prs-eth/PaGeR" target="_blank" rel="noopener noreferrer" style="display: inline-block;"> <img src="https://img.shields.io/github/stars/prs-eth/PaGeR?label=GitHub%20%E2%98%85&logo=github&color=C8C" alt="Github"> </a> <a title="Website" href="https://pager360.github.io/" target="_blank" rel="noopener noreferrer" style="display: inline-block;"> <img src="https://img.shields.io/badge/%E2%99%A5%20Project%20-Website-blue" alt="Website"> </a> <a title="arXiv" href="https://arxiv.org/abs/2605.26368" target="_blank" rel="noopener noreferrer" style="display: inline-block;"> <img src="https://img.shields.io/badge/%F0%9F%93%84%20Read%20-Paper-AF3436" alt="arXiv"> </a> <a title="Hugging Face Collection" href="https://huggingface.co/collections/prs-eth/pager-697241d06b3733a6f18e4d39" target="_blank" rel="noopener noreferrer" style="display: inline-block;"> <img src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Collection-FFD21E" alt="Hugging Face Collection"> </a> <a title="License" href="https://creativecommons.org/licenses/by/4.0/" target="_blank" rel="noopener noreferrer" style="display: inline-block;"> <img src="https://img.shields.io/badge/License-CC_BY_4.0-blue.svg" alt="License"> </a> </p> **ZüriPano** is a real-world outdoor panoramic depth benchmark, captured with the [Leica RTC360](https://leica-geosystems.com/products/laser-scanners/scanners/leica-rtc360) LiDAR scanner (8K capture, 130 m effective range, HDR + automated double-scan for transient-occlusion removal). It contains 100 equirectangular panoramas across 11 urban locations in Zürich, each paired with a dense metric depth map and a validity mask. It is used as the outdoor evaluation benchmark for [PaGeR](https://pager360.github.io/). ## Dataset Summary - **Content:** 100 outdoor scans across 11 Zürich locations, evaluation only. - **Modality:** RGB (JPG), Depth (16-bit PNG, meters via scale factor), Validity Mask (8-bit PNG), Depth Viz (8-bit Spectral RGB PNG, preview only). - **Resolution:** 4096 × 2048 equirectangular (ERP). - **Use Case:** Evaluating long-range outdoor panoramic depth estimation. ## Data Structure A single `test` split with **100 rows**, one per panorama. Each row carries: | Column | Type | Description | | :--- | :--- | :--- | | `id` | `string` | Sample id (`<Location>- s<NNN>`). | | `rgb` | `Image` | 8-bit equirectangular RGB (4096 × 2048, JPG-encoded). | | `depth` | `Image` | 16-bit single-channel PNG, `(2048, 4096)`. Decode to **meters** as `np.asarray(img, dtype=np.float32) * (200.0 / 65535.0)`. Invalid pixels are `0.0`. | | `depth_viz` | `Image` | 8-bit RGB PNG, Spectral-colormapped log-depth (per-sample min/max stretch, median-filtered). **Preview only — do NOT use for metrics or training; decode `depth` instead.** | | `mask` | `Image` | 8-bit single-channel PNG, `(2048, 4096)`. Decode as `np.asarray(img, dtype=bool)` (`255` → `True`). `True` = reliable pixel; `False` = sky, no-return, or specular surface (glass façades). Always apply when computing depth metrics. | ## How to Use ```python import numpy as np from datasets import load_dataset ds = load_dataset("prs-eth/ZuriPano", split="test") sample = ds[0] rgb = sample["rgb"] # PIL.Image, (W=4096, H=2048) depth = np.asarray(sample["depth"], dtype=np.float32) * (200.0 / 65535.0) # (2048, 4096) float32, meters mask = np.asarray(sample["mask"], dtype=bool) # (2048, 4096) bool # Always apply the mask before computing depth metrics valid_depth = depth[mask] ``` ## License ZüriPano is released under the [Creative Commons Attribution 4.0 International License (CC BY-4.0)](https://creativecommons.org/licenses/by/4.0/). You are free to share and adapt it for any purpose, including commercial use, as long as you attribute the PaGeR Authors and the ZüriPano dataset. ## Citation If you use ZüriPano in your work, please cite the PaGeR paper: ```bibtex @article{bozic2026pager, title = {Unified Panoramic Geometry Estimation via Multi-View Foundation Models}, author = {Bozic, Vukasin and Slavkovic, Isidora and Narnhofer, Dominik and Metzger, Nando and Rozumny, Denis and Schindler, Konrad and Kalischek, Nikolai}, journal = {arXiv preprint arXiv:2605.26368}, year = {2026} } ```