@surajhanchinal@YeheLiu 12M means that there are a lot of images. In this case, I suggest you to just set an upper bound on the total number of tracks (controlled by some flag which is available in config) and give it a try again.
@anuragxel@ahojnnes@mapo1 We didn't test this variant, but results on ETH3D and IMC suggest SP/ALIKED + LG isn't necessarily more accurate (but they are more robust).
What if you could combine the precision of classical Structure-from-Motion with the local robustness of feedforward 3D models?
Meet GLUEMAP — a CVPR 2026 highlight ✨ SOTA across 5 datasets spanning 6 challenges.
w/ @ahojnnes & @mapo1
🌐website: https://t.co/pEh8jQh1ZN
@anuragxel@ahojnnes@mapo1 Thanks for your questions. a) Yes—with abundant features, SIFT still gives the best accuracy (see ablation in supp. mat.). b) Nice suggestion. We followed GLOMAP's protocol and target a more general unordered setup, but it's worth exploring in future work.
📢 Introducing GenSfM – a generic Structure-from-Motion pipeline built on COLMAP, with smooth, invertible, and universal calibration for diverse cameras — from pinhole to catadioptric! Thanks @don2qian@visionviktor@mapo1 for the great collaboration.
🔗 https://t.co/Wa30YDf4vb
📢 Introducing GenSfM – a generic Structure-from-Motion pipeline built on COLMAP, with smooth, invertible, and universal calibration for diverse cameras — from pinhole to catadioptric! Thanks @don2qian@visionviktor@mapo1 for the great collaboration.
🔗 https://t.co/Wa30YDf4vb
🌺 Excited to announce the 1st CroCoDL Workshop on Large-Scale Cross-Device Localization at #ICCV2025!
Tackling real-world localization across smartphones, AR/VR, and robots with invited talks, paper track & competition on a novel dataset.
🔗 https://t.co/E41dLsdFvl
🧵 1/2
The Call for Papers is now open for our #RSS2025 workshop: Unifying Visual SLAM: From Fragmented Datasets to Scalable, Real-World Solutions
If you're working on #SLAM or #localization, we’d love to see your contributions!
https://t.co/7iA9EEMYhx
@ducha_aiki Hi, are you using Windows? Somehow windows can give this strange bug and we still haven’t figured out the reason😥. You can try limiting the maximum track number to potentially solve it (setting TrackEstablishment.max_num_tracks < 50k or so)
@pablovelagomez1@majti89@mapo1@pesarlin Hi, thanks for the effort. Do you mind sharing some of the mentioned failure cases of GLOMAP? Just for us to improve the library
🚨 The CVG group at ETH Zurich is hiring! We are looking for PhD & PostDoc in egocentric vision and video understanding.
🔗 Apply by sending an e-mail to [email protected] and become part of our amazing team! @cvg_ethz@mapo1
@pablovelagomez1@pesarlin@behnam_iut@janusch_patas No problem. I guess the gain for the global positioning would be similar to bundle adjustment but I would need to perform some tests to make an accurate conclusion there 😀
@pesarlin@pablovelagomez1@behnam_iut@janusch_patas Yes, GLOMAP does not support GPU acceleration yet. The reason for re-implementing BA was the difference in the base classes. Global positioning is slightly different as it has many extra parameters. I guess both can potentially benefit from GPU.