Weatherproofing Retrieval for Localization with Generative AI & Geometric Consistency
Yannis Kalantidis*, Mert Bulent Sariyildiz*, Rafael S. Rezende, Philippe Weinzaepfel, Diane Larlus, Gabriela Csurka
ICLR 2024
* Equal contribution
Relative gains in localization accuracy
Compared to the state of the art (black dot), using our Ret4Loc models trained with real (Ret4Loc) or real+synthetic images (Ret4Loc+Synth).
Synthetic variants
For the training set images shown on the left.
Geometric correspondences
Summary
Visual Localization Results
Place Recognition Results
Pretrained Models
Here you can find links to Ret4Loc pretrained models. We built our codebase on top of the HOW codebase. You can use code from HOW to load and evaluate the Ret4Loc models.
We provide two model weights (33MB each):
ret4loc_how.pth
– the baseline Ret4Loc-HOW modelret4loc_how_synth-pp.pth
– our best Ret4Loc model trained with synthetic data and geometric verification.
You can load our models exactly like a HOW model.