ProBA: Probabilistic Bundle Adjustment with the Bhattacharyya Coefficient
This work addresses a key limitation in SLAM systems for unstructured environments by making BA more robust and initialization-free, though it is an incremental improvement over existing probabilistic methods.
The paper tackles the problem of Bundle Adjustment (BA) requiring accurate initial estimates and known camera intrinsics by proposing ProBA, a probabilistic formulation that models uncertainty in 2D observations and 3D structure, enabling optimization without prior knowledge. Experimental results show ProBA outperforms traditional methods in challenging real-world conditions, enhancing the practicality of SLAM systems in unstructured environments.
Classical Bundle Adjustment (BA) methods require accurate initial estimates for convergence and typically assume known camera intrinsics, which limits their applicability when such information is uncertain or unavailable. We propose a novel probabilistic formulation of BA (ProBA) that explicitly models and propagates uncertainty in both the 2D observations and the 3D scene structure, enabling optimization without any prior knowledge of camera poses or focal length. Our method uses 3D Gaussians instead of point-like landmarks and we introduce uncertainty-aware reprojection losses by projecting the 3D Gaussians onto the 2D image space, and enforce geometric consistency across multiple 3D Gaussians using the Bhattacharyya coefficient to encourage overlap between their corresponding Gaussian distributions. This probabilistic framework leads to more robust and reliable optimization, even in the presence of outliers in the correspondence set, reducing the likelihood of converging to poor local minima. Experimental results show that \textit{ProBA} outperforms traditional methods in challenging real-world conditions. By removing the need for strong initialization and known intrinsics, ProBA enhances the practicality of SLAM systems deployed in unstructured environments.