Hierarchical Visual Relocalization with Nearest View Synthesis from Feature Gaussian Splatting
This work addresses a fundamental challenge in 3D computer vision for applications like robotics and augmented reality, offering incremental improvements in scalability and accuracy.
The paper tackles the problem of visual relocalization, where camera pose estimation is limited by sparse image observations and weak feature matching, by proposing SplatHLoc, a hierarchical framework using Feature Gaussian Splatting, adaptive viewpoint synthesis, and hybrid feature matching, resulting in enhanced robustness and new state-of-the-art performance on indoor and outdoor datasets.
Visual relocalization is a fundamental task in the field of 3D computer vision, estimating a camera's pose when it revisits a previously known scene. While point-based hierarchical relocalization methods have shown strong scalability and efficiency, they are often limited by sparse image observations and weak feature matching. In this work, we propose SplatHLoc, a novel hierarchical visual relocalization framework that uses Feature Gaussian Splatting as the scene representation. To address the sparsity of database images, we propose an adaptive viewpoint retrieval method that synthesizes virtual candidates with viewpoints more closely aligned with the query, thereby improving the accuracy of initial pose estimation. For feature matching, we observe that Gaussian-rendered features and those extracted directly from images exhibit different strengths across the two-stage matching process: the former performs better in the coarse stage, while the latter proves more effective in the fine stage. Therefore, we introduce a hybrid feature matching strategy, enabling more accurate and efficient pose estimation. Extensive experiments on both indoor and outdoor datasets show that SplatHLoc enhances the robustness of visual relocalization, setting a new state-of-the-art.