Uploaded July 2026 | Updated November 2026, 46 minutes ago
Reliable odometry is essential for mobile robots as they increasingly enter more challenging environments, which often contain little information to constrain point cloud registration, resulting in degraded LiDAR-Inertial Odometry (LIO) accuracy or even divergence. To address this, we present BIEVR-LIO, a novel approach designed specifically to exploit subtle variations in the available geometry for improved robustness. We propose a high-resolution map representation that stores surfaces as voxel-wise oriented height images. This representation can directly be used for registration without the calculation of intermediate geometric primitives while still supporting efficient updates. Since informative geometry is often sparsely distributed in the environment, we further propose a map-informed point sampling strategy to focus registration on geometrically informative regions, improving robustness in uninformative environments while reducing computational cost compared to global high-resolution sampling. Experiments across multiple sensors, platforms, and environments demonstrate state-of-the-art performance in well-constrained scenes and substantial improvements in challenging scenarios where baseline methods diverge. Additionally, we demonstrate that the fine-grained geometry captured by BIEVR-LIO can be used for downstream tasks such as elevation mapping for robot locomotion.
Patrick Pfreundschuh, Turcan Tuna, Cedric Le Gentil, Roland Siegwart, Cesar Cadena, Helen Oleynikova
Project Page: patripfr.github.io/bievr-lio
Paper: arxiv.org/abs/2604.14421
Code: github.com/ethz-asl/BIEVR-LIO
Reliable odometry is essential for mobile robots as they increasingly enter more challenging environments, which often contain little information to constrain point cloud registration, resulting in degraded LiDAR-Inertial Odometry (LIO) accuracy or even divergence. To address this, we present BIEVR-LIO, a novel approach designed specifically to exploit subtle variations in the available geometry for improved robustness. We propose a high-resolution map representation that stores surfaces as voxel-wise oriented height images. This representation can directly be used for registration without the calculation of intermediate geometric primitives while still supporting efficient updates. Since informative geometry is often sparsely distributed in the environment, we further propose a map-informed point sampling strategy to focus registration on geometrically informative regions, improving robustness in uninformative environments while reducing computational cost compared to global high-resolution sampling. Experiments across multiple sensors, platforms, and environments demonstrate state-of-the-art performance in well-constrained scenes and substantial improvements in challenging scenarios where baseline methods diverge. Additionally, we demonstrate that the fine-grained geometry captured by BIEVR-LIO can be used for downstream tasks such as elevation mapping for robot locomotion.
Patrick Pfreundschuh, Turcan Tuna, Cedric Le Gentil, Roland Siegwart, Cesar Cadena, Helen Oleynikova
Project Page: patripfr.github.io/bievr-lio
Paper: arxiv.org/abs/2604.14421
Code: github.com/ethz-asl/BIEVR-LIO








![[CVPR 2023] Unsupervised Continual Semantic Adaptation through Neural Rendering
This is the accompanying video of our CVPR 2023 paper Unsupervised Continual Semantic Adaptation through Neural Rendering.
Paper: https://arxiv.org/abs/2211.13969
Project page: https://ethz-asl.github.io/ucsa_neural_rendering
Code: https://github.com/ethz-asl/ucsa_neural_rendering
Abstract:
An increasing amount of applications rely on data-driven models that are deployed for perception tasks across a sequence of scenes. Due to the mismatch between training and deployment data, adapting the model on the new scenes is often crucial to obtain good performance. In this work, we study continual multi-scene adaptation for the task of semantic segmentation, assuming that no ground-truth labels are available during deployment and that performance on the previous scenes should be maintained. We propose training a Semantic-NeRF network for each scene by fusing the predictions of a segmentation model and then using the view-consistent rendered semantic labels as pseudo-labels to adapt the model. Through joint training with the segmentation model, the Semantic-NeRF model effectively enables 2D-3D knowledge transfer. Furthermore, due to its compact size, it can be stored in a long-term memory and subsequently used to render data from arbitrary viewpoints to reduce forgetting. We evaluate our approach on ScanNet, where we outperform both a voxel-based baseline and a state-of-the-art unsupervised domain adaptation method.
Reference:
Zhizheng Liu*, Francesco Milano*, Jonas Frey, Roland Siegwart, Hermann Blum and Cesar Cadena, Unsupervised Continual Semantic Adaptation through Neural Rendering, IEEE/CVF Computer Vision and Pattern Recognition Conference (CVPR), 2023.
Affiliations:
Autonomous Systems Laboratory, ETH Zurich, Switzerland.
Robotic Systems Laboratory, ETH Zurich, Switzerland.
Max Planck ETH Center for Learning Systems, Switzerland. [CVPR 2023] Unsupervised Continual Semantic Adaptation through Neural Rendering](https://i.ytimg.com/vi/XfNLsl8ATNY/mqdefault.jpg)

