# ACE-G: Improving Generalization of Scene Coordinate Regression Through Query Pre-Training arXiv: https://arxiv.org/abs/2510.11605 ยท HTML: https://mlsanity.com/paper/2510.11605/ ## Key facts - Cited by 9 papers in the map (data through 14 Oct 2026). - Builds on 53 of its 70 references that are in the map. - First posted on arXiv 13 Oct 2025. ## Builds on (most cited first, 30 of 53) - [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020) - [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023) - [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023) - [Layer Normalization](https://mlsanity.com/paper/1607.06450/) (2016) - [ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes](https://mlsanity.com/paper/1702.04405/) (2017) - [DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation](https://mlsanity.com/paper/1901.05103/) (2019) - [DUSt3R: Geometric 3D Vision Made Easy](https://mlsanity.com/paper/2312.14132/) (2023) - [SuperGlue: Learning Feature Matching with Graph Neural Networks](https://mlsanity.com/paper/1911.11763/) (2019) - [SuperPoint: Self-Supervised Interest Point Detection and Description](https://mlsanity.com/paper/1712.07629/) (2017) - [LoFTR: Detector-Free Local Feature Matching with Transformers](https://mlsanity.com/paper/2104.00680/) (2021) - [Grounding Image Matching in 3D with MASt3R](https://mlsanity.com/paper/2406.09756/) (2024) - [Vision Transformers Need Registers](https://mlsanity.com/paper/2309.16588/) (2023) - [ScanNet++: A High-Fidelity Dataset of 3D Indoor Scenes](https://mlsanity.com/paper/2308.11417/) (2023) - [From Coarse to Fine: Robust Hierarchical Localization at Large Scale](https://mlsanity.com/paper/1812.03506/) (2018) - [BlendedMVS: A Large-scale Dataset for Generalized Multi-view Stereo Networks](https://mlsanity.com/paper/1911.10127/) (2019) - [DSAC - Differentiable RANSAC for Camera Localization](https://mlsanity.com/paper/1611.05705/) (2016) - [Image Matching across Wide Baselines: From Paper to Practice](https://mlsanity.com/paper/2003.01587/) (2020) - [Back to the Feature: Learning Robust Camera Localization from Pixels to Pose](https://mlsanity.com/paper/2103.09213/) (2021) - [Understanding the Limitations of CNN-based Absolute Camera Pose Regression](https://mlsanity.com/paper/1903.07504/) (2019) - [Visual Camera Re-Localization from RGB and RGB-D Images Using DSAC](https://mlsanity.com/paper/2002.12324/) (2020) - [Optimizing the Latent Space of Generative Networks](https://mlsanity.com/paper/1707.05776/) (2017) - [Learning Less is More - 6D Camera Localization via 3D Surface Regression](https://mlsanity.com/paper/1711.10228/) (2017) - [LENS: Localization enhanced by NeRF synthesis](https://mlsanity.com/paper/2110.06558/) (2021) - [RGBD Objects in the Wild: Scaling Real-World 3D Object Learning from RGB-D Videos](https://mlsanity.com/paper/2401.12592/) (2024) - [Map-free Visual Relocalization: Metric Pose Relative to a Single Image](https://mlsanity.com/paper/2210.05494/) (2022) - [Learning Multi-Scene Absolute Pose Regression with Transformers](https://mlsanity.com/paper/2103.11468/) (2021) - [Accelerated Coordinate Encoding: Learning to Relocalize in Minutes using RGB and Poses](https://mlsanity.com/paper/2305.14059/) (2023) - [Hierarchical Scene Coordinate Classification and Regression for Visual Localization](https://mlsanity.com/paper/1909.06216/) (2019) - [AtLoc: Attention Guided Camera Localization](https://mlsanity.com/paper/1909.03557/) (2019) - [DFNet: Enhance Absolute Pose Regression with Direct Feature Matching](https://mlsanity.com/paper/2204.00559/) (2022) ## Built on by (most cited first, 9 of 9) - [ULF-Loc: Unbiased Landmark Feature for Robust Visual Localization with 3D Gaussian Splatting](https://mlsanity.com/paper/2605.04730/) (2026) - [ACE-SLAM: Scene Coordinate Regression for Neural Implicit Real-Time SLAM](https://mlsanity.com/paper/2512.14032/) (2025) - [MultiLoc: Look Around As You Localize for Fast And Robust Visual Re-localization](https://mlsanity.com/paper/2603.27170/) (2026) - [Hierarchical Visual Relocalization with Nearest View Synthesis from Feature Gaussian Splatting](https://mlsanity.com/paper/2603.29185/) (2026) - [Seeing Through the Weights: Privacy Leakage in Scene Coordinate Regression](https://mlsanity.com/paper/2606.31164/) (2026) - [LightLoc++: Sensor-Robust Representation Learning for Efficient Outdoor LiDAR Localization](https://mlsanity.com/paper/2608.15317/) (2026) - [AutoCompass: Accurate Visual Localization on Public Maps by Learning from Weak Labels](https://mlsanity.com/paper/2609.02798/) (2026) - [BLASt3R: Bundle Adjustment of Any Image Set with Multi-View Matching and Monocular Priors](https://mlsanity.com/paper/2609.05210/) (2026) - [VideoReloc: Long-Term Indoor Video Relocalization against a Kilobyte-Scale Semantic Scene Graph](https://mlsanity.com/paper/2609.21804/) (2026) ## Abstract (arXiv) Scene coordinate regression (SCR) has established itself as a promising learning-based approach to visual relocalization. After mere minutes of scene-specific training, SCR models estimate camera poses of query images with high accuracy. Still, SCR methods fall short of the generalization capabilities of more classical feature-matching approaches. When imaging conditions of query images, such as lighting or viewpoint, are too different from the training views, SCR models fail. Failing to generalize is an inherent limitation of previous SCR frameworks, since their training objective is to encode the training views in the weights of the coordinate regressor itself. The regressor essentially overfits to the training views, by design. We propose to separate the coordinate regressor and the map representation into a generic transformer and a scene-specific map code. This separation allows us to pre-train the transformer on tens of thousands of scenes. More importantly, it allows us to train the transformer to generalize from mapping images to unseen query images during pre-training. We demonstrate on multiple challenging relocalization datasets that our method, ACE-G, leads to significantly increased robustness while keeping the computational footprint attractive.