Con-DSO: Learning Short-Horizon Consistency Priors for RGB-D Direct Sparse Odometry

• URL: http://arxiv.org/abs/2605.27952v1
• Date: Wed, 27 May 2026 04:46:52 GMT
• Title: Con-DSO: Learning Short-Horizon Consistency Priors for RGB-D Direct Sparse Odometry
• Abstract summary: Con-DSO is a consistency-aware RGB-D direct sparse odometry framework.<n>It predicts dense photometric and depth-geometric consistency uncertainty from temporally adjacent RGB-D frame pairs.<n>Experiments show substantial gains over direct RGB-D VO baselines.
• Score: 8.079139316299182
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Visual odometry (VO) is a fundamental component in robotics and augmented reality. RGB-D direct VO benefits from metric depth measurements, but it can degrade in challenging environments, where dynamic objects, occlusions, illumination changes, and unreliable depth violate the short-horizon photometric and depth-geometric consistency assumptions used by direct alignment. Existing approaches mitigate these issues through semantic filtering, explicit occlusion reasoning, illumination adaptation, or hand-crafted geometric criteria, but often rely on external modules or fixed assumptions tailored to individual failure modes, limiting their flexibility and ability to handle diverse challenges in a unified manner. In this work, we propose Con-DSO, a consistency-aware RGB-D direct sparse odometry framework that predicts dense photometric and depth-geometric consistency uncertainty from temporally adjacent RGB-D frame pairs. The consistency network is trained using flow-guided photometric errors and projective depth-consistency errors, allowing consistency violations to be represented as pixel-level uncertainty. These pairwise uncertainty predictions are converted into a host-side quality prior for keyframe-based tracking. The prior is then applied to VO through quality-aware support-pixel selection and decoupled photometric-geometric weighting during pose estimation, enabling continuous attenuation of unreliable observations rather than hard rejection or threshold-based gating. Experiments on five public RGB-D benchmarks show substantial gains over direct RGB-D VO baselines, with over 20\% absolute trajectory error reduction on ICL-NUIM and 50\%--80\% reductions on RGB-D Scenes V2, TUM/Bonn Dynamic, and OpenLORIS sequences.

Related papers

• Lightweight Interpretable RGB-Guided Hyperspectral Super-Resolution under Real Cross-resolution Misalignment [6.5685272845309255]
  RGB-guided hyperspectral super-resolution transfers spatial detail from a high-resolution RGB guide to a low-resolution hyperspectral image.<n>We propose a lightweight and interpretable RGB-guided HSR framework combining cross-modal flow alignment with model-based Gram-Schmidtization fusion.
  arXiv  Detail & Related papers  (2026-09-01T10:55:11Z)
• Cyclops: LiDAR as a Camera That Dreams in Color [32.25721672776894]
  We propose a framework that translates sparse Non-Repetitive Scanning LiDAR (NRS-LiDAR) intensity into RGB video, enabling camera-free inference for all-day perception tasks.<n>Our approach first converts sparse LiDAR intensity projections into dense representations via a frozen pre-trained densification module.<n>The dense intensity latent is then transported toward the target RGB distribution through Latent Bridge Matching.
  arXiv  Detail & Related papers  (2026-08-17T08:37:09Z)
• BRIC-Net: Boundary-Reliable Illumination-Color Interaction for Remote Sensing Image Deshadowing [7.326663156595675]
  Shadows in remote sensing images obscure surface appearance and disrupt radiometric continuity.<n>We propose the Boundary-Reliable Illumination-Color Interaction Network (BRIC-Net), which decouples these failures at different representation levels.<n>BRIC-Net achieves 29.46dB full-image peak signal-to-noise ratio (PSNR) on AeroDS-Syn and 27.96dB on SRGTA.
  arXiv  Detail & Related papers  (2026-08-01T14:07:43Z)
• On the Global Photometric Alignment for Low-Level Vision [14.897162428029878]
  Supervised low-level vision models rely on pixel-wise losses against paired references, yet paired training sets exhibit per-pair photometric inconsistency.<n>Standard reconstruction losses allocate disproportionate gradient budget to conflicting per-pair photometric targets, crowding out content restoration.<n>Motivated by this analysis, we propose Photometric Alignment Loss (PAL)
  arXiv  Detail & Related papers  (2026-04-09T12:29:54Z)
• Learning Latent Proxies for Controllable Single-Image Relighting [29.76860042350802]
  Single-image relighting is highly under-constrained.<n>We introduce LightCtrl that integrates physical priors at two levels.<n>We also present ScaLight, a large-scale object-level dataset.
  arXiv  Detail & Related papers  (2026-03-16T17:16:59Z)
• Depth-Guided Metric-Aware Temporal Consistency for Monocular Video Human Mesh Recovery [28.9705779052755]
  We propose a comprehensive framework that achieves metric-aware temporal consistency through three synergistic components.<n>A Depth-Guided Multi-Scale Fusion module adaptively integrates geometric priors with RGB features via confidence-aware gating.<n>A Motion-Depth Aligned Refinement (MoDAR) module that enforces temporal coherence through cross-modal attention between motion dynamics and geometric cues.
  arXiv  Detail & Related papers  (2026-02-04T06:41:03Z)
• HAD: Hierarchical Asymmetric Distillation to Bridge Spatio-Temporal Gaps in Event-Based Object Tracking [80.07224739976911]
  Event cameras offer exceptional temporal resolution and a range (modal)<n> RGB cameras excel at capturing rich texture with high resolution, whereas event cameras offer exceptional temporal resolution and a range (modal)
  arXiv  Detail & Related papers  (2025-10-22T13:15:13Z)
• SplatPose: Geometry-Aware 6-DoF Pose Estimation from Single RGB Image via 3D Gaussian Splatting [3.6688867031495223]
  We introduce SplatPose, a novel framework that synergizes 3D Gaussian Splatting (3DGS) with a dual-branch neural architecture to achieve high-precision pose estimation.<n> Experiments on three benchmark datasets demonstrate that SplatPose achieves state-of-the-art 6-DoF pose estimation accuracy in single RGB settings.
  arXiv  Detail & Related papers  (2025-03-07T06:40:06Z)
• Generalizable Non-Line-of-Sight Imaging with Learnable Physical Priors [52.195637608631955]
  Non-line-of-sight (NLOS) imaging has attracted increasing attention due to its potential applications. Existing NLOS reconstruction approaches are constrained by the reliance on empirical physical priors. We introduce a novel learning-based solution, comprising two key designs: Learnable Path Compensation (LPC) and Adaptive Phasor Field (APF)
  arXiv  Detail & Related papers  (2024-09-21T04:39:45Z)
• RGB-based Category-level Object Pose Estimation via Decoupled Metric Scale Recovery [72.13154206106259]
  We propose a novel pipeline that decouples the 6D pose and size estimation to mitigate the influence of imperfect scales on rigid transformations. Specifically, we leverage a pre-trained monocular estimator to extract local geometric information. A separate branch is designed to directly recover the metric scale of the object based on category-level statistics.
  arXiv  Detail & Related papers  (2023-09-19T02:20:26Z)
• Optimal baseline exploitation in vertical dark-matter detectors based on atom interferometry [50.06952271801328]
  Several terrestrial detectors for gravitational waves and dark matter based on long-baseline atom interferometry are currently in the final planning stages or already under construction. We show that resonant-mode detectors based on multi-diamond fountain gradiometers achieve the optimal, shot-noise limited, sensitivity if their height constitutes 20% of the available baseline.
  arXiv  Detail & Related papers  (2023-09-08T08:38:24Z)
• Polarimetric Information for Multi-Modal 6D Pose Estimation of Photometrically Challenging Objects with Limited Data [51.95347650131366]
  6D pose estimation pipelines that rely on RGB-only or RGB-D data show limitations for photometrically challenging objects. A supervised learning-based method utilising complementary polarisation information is proposed to overcome such limitations.
  arXiv  Detail & Related papers  (2023-08-21T10:56:00Z)
• Symmetric Uncertainty-Aware Feature Transmission for Depth Super-Resolution [52.582632746409665]
  We propose a novel Symmetric Uncertainty-aware Feature Transmission (SUFT) for color-guided DSR. Our method achieves superior performance compared to state-of-the-art methods.
  arXiv  Detail & Related papers  (2023-06-01T06:35:59Z)
• Frequency-Aware Self-Supervised Monocular Depth Estimation [41.97188738587212]
  We present two versatile methods to enhance self-supervised monocular depth estimation models. The high generalizability of our methods is achieved by solving the fundamental and ubiquitous problems in photometric loss function. We are the first to propose blurring images to improve depth estimators with an interpretable analysis.
  arXiv  Detail & Related papers  (2022-10-11T14:30:26Z)

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.