Visual Odometry with Transformers

• URL: http://arxiv.org/abs/2510.03348v1
• Date: Thu, 02 Oct 2025 17:00:14 GMT
• Title: Visual Odometry with Transformers
• Authors: Vlardimir Yugay, Duy-Kien Nguyen, Theo Gevers, Cees G. M. Snoek, Martin R. Oswald,
• Abstract summary: We introduce Visual odometry Transformer (VoT), which processes sequences of monocular frames by extracting features.<n>Unlike prior methods, VoT directly predicts camera motion without estimating dense geometry and relies solely on camera poses for supervision.<n>VoT scales effectively with larger datasets, benefits substantially from stronger pre-trained backbones, generalizes across diverse camera motions and calibration settings, and outperforms traditional methods while running more than 3 times faster.
• Score: 68.453547770334
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Modern monocular visual odometry methods typically combine pre-trained deep learning components with optimization modules, resulting in complex pipelines that rely heavily on camera calibration and hyperparameter tuning, and often struggle in unseen real-world scenarios. Recent large-scale 3D models trained on massive amounts of multi-modal data have partially alleviated these challenges, providing generalizable dense reconstruction and camera pose estimation. Still, they remain limited in handling long videos and providing accurate per-frame estimates, which are required for visual odometry. In this work, we demonstrate that monocular visual odometry can be addressed effectively in an end-to-end manner, thereby eliminating the need for handcrafted components such as bundle adjustment, feature matching, camera calibration, or dense 3D reconstruction. We introduce VoT, short for Visual odometry Transformer, which processes sequences of monocular frames by extracting features and modeling global relationships through temporal and spatial attention. Unlike prior methods, VoT directly predicts camera motion without estimating dense geometry and relies solely on camera poses for supervision. The framework is modular and flexible, allowing seamless integration of various pre-trained encoders as feature extractors. Experimental results demonstrate that VoT scales effectively with larger datasets, benefits substantially from stronger pre-trained backbones, generalizes across diverse camera motions and calibration settings, and outperforms traditional methods while running more than 3 times faster. The code will be released.

Related papers

• Stereo-Inertial Poser: Towards Metric-Accurate Shape-Aware Motion Capture Using Sparse IMUs and a Single Stereo Camera [54.967647497048205]
  We present Stereo-Inertial Poser, a real-time motion capture system that estimates metric-accurate and shape-aware 3D human motion.<n>We replace the monocular RGB with stereo vision, enabling direct 3D keypoint extraction and body shape parameter estimation.<n>Our method produces drift-free global translation under a long recording time and reduces foot-skating effects.
  arXiv  Detail & Related papers  (2026-03-02T17:46:38Z)
• SV-GS: Sparse View 4D Reconstruction with Skeleton-Driven Gaussian Splatting [19.12278036176021]
  We present SV-GS, a framework that simultaneously estimates a deformation model and the object's motion over time under sparse observations.<n>Our method outperforms existing approaches under sparse observations by up to 34% in PSNR.
  arXiv  Detail & Related papers  (2026-01-01T09:53:03Z)
• MapAnything: Universal Feed-Forward Metric 3D Reconstruction [63.79151976126576]
  MapAnything ingests one or more images along with optional geometric inputs such as camera intrinsics, poses, depth, or partial reconstructions.<n>It then directly regresses the metric 3D scene geometry and cameras.<n>MapAnything addresses a broad range of 3D vision tasks in a single feed-forward pass.
  arXiv  Detail & Related papers  (2025-09-16T18:00:14Z)
• Multi-LVI-SAM: A Robust LiDAR-Visual-Inertial Odometry for Multiple Fisheye Cameras [24.95534498798919]
  We propose a multi-camera LiDAR-visual-inertial odometry framework, Multi-LVI-SAM, which fuses data from multiple fisheye cameras, LiDAR and inertial sensors for highly accurate and robust state estimation.<n>To enable efficient and consistent integration of visual information from multiple fisheye cameras, we introduce a panoramic visual feature model that unifies multi-camera observations into a single representation.
  arXiv  Detail & Related papers  (2025-09-06T15:06:55Z)
• DGS-LRM: Real-Time Deformable 3D Gaussian Reconstruction From Monocular Videos [52.46386528202226]
  We introduce the Deformable Gaussian Splats Large Reconstruction Model (DGS-LRM)<n>It is the first feed-forward method predicting deformable 3D Gaussian splats from a monocular posed video of any dynamic scene.<n>It achieves performance on par with state-of-the-art monocular video 3D tracking methods.
  arXiv  Detail & Related papers  (2025-06-11T17:59:58Z)
• SpatialCrafter: Unleashing the Imagination of Video Diffusion Models for Scene Reconstruction from Limited Observations [44.53106180688135]
  This work takes on the challenge of reconstructing 3D scenes from sparse or single-view inputs.<n>We introduce SpatialCrafter, a framework that leverages the rich knowledge in video diffusion models to generate plausible additional observations.<n>Through a trainable camera encoder and an epipolar attention mechanism for explicit geometric constraints, we achieve precise camera control and 3D consistency.
  arXiv  Detail & Related papers  (2025-05-17T13:05:13Z)
• ESVO2: Direct Visual-Inertial Odometry with Stereo Event Cameras [41.992980062962495]
  Event-based visual odometry aims at solving tracking and mapping subproblems (typically in parallel)<n>We build an event-based stereo visual-inertial odometry system on top of a direct pipeline.<n>The resulting system scales well with modern high-resolution event cameras.
  arXiv  Detail & Related papers  (2024-10-12T05:35:27Z)
• MultiViPerFrOG: A Globally Optimized Multi-Viewpoint Perception Framework for Camera Motion and Tissue Deformation [18.261678529996104]
  We propose a framework that can flexibly integrate the output of low-level perception modules with kinematic and scene-modeling priors. Overall, our method shows robustness to combined noisy input measures and can process hundreds of points in a few milliseconds.
  arXiv  Detail & Related papers  (2024-08-08T10:55:55Z)
• SceNeRFlow: Time-Consistent Reconstruction of General Dynamic Scenes [75.9110646062442]
  We propose SceNeRFlow to reconstruct a general, non-rigid scene in a time-consistent manner. Our method takes multi-view RGB videos and background images from static cameras with known camera parameters as input. We show experimentally that, unlike prior work that only handles small motion, our method enables the reconstruction of studio-scale motions.
  arXiv  Detail & Related papers  (2023-08-16T09:50:35Z)
• FrozenRecon: Pose-free 3D Scene Reconstruction with Frozen Depth Models [67.96827539201071]
  We propose a novel test-time optimization approach for 3D scene reconstruction. Our method achieves state-of-the-art cross-dataset reconstruction on five zero-shot testing datasets.
  arXiv  Detail & Related papers  (2023-08-10T17:55:02Z)
• Towards Scalable Multi-View Reconstruction of Geometry and Materials [27.660389147094715]
  We propose a novel method for joint recovery of camera pose, object geometry and spatially-varying Bidirectional Reflectance Distribution Function (svBRDF) of 3D scenes. The input are high-resolution RGBD images captured by a mobile, hand-held capture system with point lights for active illumination.
  arXiv  Detail & Related papers  (2023-06-06T15:07:39Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.