Hierarchical Localization Versions Save

• New front ends
  • global features: OpenIBL (https://github.com/cvg/Hierarchical-Localization/pull/164), CosPlace (https://github.com/cvg/Hierarchical-Localization/pull/257 by @gmberton)
  • patch descriptors: SOSNet (https://github.com/cvg/Hierarchical-Localization/pull/161 by @mihaidusmanu), HardNet (https://github.com/cvg/Hierarchical-Localization/pull/235)
  • detector & descriptor: DISK (https://github.com/cvg/Hierarchical-Localization/pull/233 by @lxxue, https://github.com/cvg/Hierarchical-Localization/pull/291)
  • sparse matching: AdaLAM (https://github.com/cvg/Hierarchical-Localization/pull/229 by @mihaidusmanu), LightGlue (https://github.com/cvg/Hierarchical-Localization/pull/285 by @Phil26AT)
  • dense matching: LoFTR (https://github.com/cvg/Hierarchical-Localization/pull/173, https://github.com/cvg/Hierarchical-Localization/pull/243, https://github.com/cvg/Hierarchical-Localization/pull/254 by @Phil26AT)
• Triangulation: use known camera poses for two-view geometric verification (https://github.com/cvg/Hierarchical-Localization/pull/178 by @mihaidusmanu)
• Control over COLMAP import and reconstruction options (https://github.com/cvg/Hierarchical-Localization/pull/210)
• Performance
  • More reliably skip existing pairs in a match file (https://github.com/cvg/Hierarchical-Localization/pull/159)
  • Faster HDF5 write (https://github.com/cvg/Hierarchical-Localization/pull/194 by @clementinboittiaux)
  • Parallel reading and writing in match_features (https://github.com/cvg/Hierarchical-Localization/pull/242)
• Add scalar detection uncertainty for LaMAR (https://github.com/cvg/Hierarchical-Localization/pull/158)
• Documentation (https://github.com/cvg/Hierarchical-Localization/pull/294)
• Updated requirements: tqdm>=4.36.0, pycolmap>=0.3.0, kornia>=0.6.11

Thanks to all contributors: @mihaidusmanu @Phil26AT @clementinboittiaux @gmberton @lxxue @soskek @mgprt @lasuomela @francescomilano172 @pyetras @Keunmo @jiaqchen

This is an important release that makes hloc easier to use and more efficient on many points. The main win is that we now can skip the painful installation of COLMAP and install all dependencies with pip. This means that we can run hloc in Google Colab, with GPUs, for free. Most of these additions have been made possible by the awesome work of @Phil26AT on pycolmap.

• Demo notebook in Google Colab - SfM and localization right in your browser!
• Use the new pycolmap Reconstruction objects and pipeline API
  • Do not require an installation of COLMAP anymore - pycolmap is enough
  • Faster model reading and writing
  • Fine-grained control over camera sharing via the camera_mode parameter
  • Localization with unknown or inaccurate focal length
• Modular localization API with control over all estimator parameters
• 3D visualizations or camera frustums and points with plotly
• Package-specific logging in the hloc namespace
• Store the extracted features by default as fp16 instead of fp32
• Optionally fix a long-standing bug in SuperPoint descriptor sampling
• Add script to compute exhaustive pairs for reconstruction or localization
• Require pycolmap>=0.1.0

December 2021.

• Bug fixes and usability improvements.
• Support PIL backend for image resizing.
• Add __version__ attribute to check against future releases.

July 2021

• [BREAKING] improved structure of the SfM folders (triangulation and reconstruction), see #76
• Support for image retrieval (NetVLAD, DIR) and more local features (SIFT, R2D2)
• Support for more datasets: Aachen v1.1, Extended CMU Seasons, RobotCar Seasons, 4Seasons, Cambridge Landmarks, 7-Scenes
• Simplified pipeline and API
• Spatial matcher
• Support for arbitrary paths of features and matches
• Support for matching multiple feature files together

First public version, released in July 2020.