RePLAy: Remove Projective LiDAR Depthmap Artifacts via Exploiting Epipolar Geometry
This addresses a critical issue for autonomous vehicle systems by enabling artifact-free depthmaps in datasets lacking stereo images, though it is an incremental improvement over existing heuristic solutions.
The paper tackles the problem of systematic misalignment artifacts in LiDAR projected depthmaps for autonomous vehicles, proposing a parameter-free analytical solution called RePLAy that removes these artifacts and unanimously improves state-of-the-art monocular depth estimators and 3D object detectors.
3D sensing is a fundamental task for Autonomous Vehicles. Its deployment often relies on aligned RGB cameras and LiDAR. Despite meticulous synchronization and calibration, systematic misalignment persists in LiDAR projected depthmap. This is due to the physical baseline distance between the two sensors. The artifact is often reflected as background LiDAR incorrectly projected onto the foreground, such as cars and pedestrians. The KITTI dataset uses stereo cameras as a heuristic solution to remove artifacts. However most AV datasets, including nuScenes, Waymo, and DDAD, lack stereo images, making the KITTI solution inapplicable. We propose RePLAy, a parameter-free analytical solution to remove the projective artifacts. We construct a binocular vision system between a hypothesized virtual LiDAR camera and the RGB camera. We then remove the projective artifacts by determining the epipolar occlusion with the proposed analytical solution. We show unanimous improvement in the State-of-The-Art (SoTA) monocular depth estimators and 3D object detectors with the artifacts-free depthmaps.