Xecades/AerialExtreMatch-Train

1---2license: mit3task_categories:4- depth-estimation5- keypoint-detection6- image-feature-extraction7pretty_name: AerialExtreMatch Train8viewer: false9tags:10- image11---12# AerialExtreMatch — Train Dataset13 14[Code](https://github.com/Xecades/AerialExtreMatch) | [Project Page](https://xecades.github.io/AerialExtreMatch/) | [Paper](https://ieeexplore.ieee.org/document/11434521)15 16This repo contains the **training** set for our paper *AerialExtreMatch: A Benchmark for Extreme-View Image Matching and Localization*. We also provide [**benchmark**](https://huggingface.co/datasets/Xecades/AerialExtreMatch-Benchmark) and [**localization**](https://huggingface.co/datasets/Xecades/AerialExtreMatch-Localization) datasets.17 18## Usage19 20Simply clone this repository and unzip the dataset files.21 22```bash23git clone git@hf.co:datasets/Xecades/AerialExtreMatch-Train24cd AerialExtreMatch-Train25unzip "*.zip"26rm -rf *.zip27rm -rf .git28```29 30## Dataset Structure31 32After unpacking each .zip file:33 34<pre>35.36├── scene_info: *.npz <i>(e.g. Tokyo3_100@-98@-185@67@0@90.npz)</i>37├── geo_info: *.txt <i>(e.g. Tokyo_3.txt, see below)</i>38└── [city_id] <i>(e.g. Tokyo3)</i>39    ├── computed_pairs.txt40    ├── overlap_5_30.txt41    ├── overlap_30_50.txt42    ├── overlap_50_plus.txt43    ├── query44    │   ├── intrinsic.txt <i>(see below)</i>45    │   ├── poses.txt <i>(see below)</i>46    │   ├── depth: *.exr47    │   └── rgb: *.jpg48    └── reference49        ├── intrinsic.txt <i>(see below)</i>50        ├── poses.txt <i>(see below)</i>51        ├── depth: *.exr52        └── rgb: *.jpg53</pre>54 55 - Keys of `.npz` files in `scene_info`: `['poses', 'intrinsics', 'depth', 'rgb', 'overlap', 'pair']`.56 - Format of `.txt` files in `geo_info`: `name lat1 lon1 lat2 lon2 alt`.57 - Format of `intrinsic.txt`: `name PINHOLE w h fx fy cx cy`.58 - Format of `poses.txt`: `name qw qx qy qz x y z`. (w2c)