Hierarchical-Localization

Why we included this project

hloc implements hierarchical localization, which pairs an image-retrieval step with feature matching to estimate a camera's 6-DoF pose against a prebuilt 3D map. The toolbox ships with pipelines that reproduce state-of-the-art results on benchmarks like Aachen and InLoc, and its modular structure makes it easy to swap in your own feature extractors and evaluate them side by side with the bundled SuperPoint and SuperGlue. That combination of working reference pipelines and interchangeable parts is what makes it useful for teams building AR or robotics systems that need to know where a camera is relative to a known map, and equally for researchers testing new components. Because it bundles several well-known algorithms behind consistent interfaces, it also serves as a practical way to learn how modern SfM and visual localization pipelines fit together.