Large Depth Completion Model from Sparse Observations

• URL: http://arxiv.org/abs/2605.30115v1
• Date: Thu, 28 May 2026 15:50:29 GMT
• Title: Large Depth Completion Model from Sparse Observations
• Abstract summary: Large Depth Completion Model (LDCM) is a simple, effective, and robust framework for single-view metric depth estimation with sparse observations.<n>LDCM generates metric-accurate dense depth maps using a transformer.
• Score: 28.01552819701954
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: This work presents the Large Depth Completion Model (LDCM), a simple, effective, and robust framework for single-view metric depth estimation with sparse observations. Without relying on complex architectural designs, LDCM generates metric-accurate dense depth maps using a transformer. It outperforms existing approaches across diverse datasets and sparse observations. We achieve this from two key perspectives: (1) leveraging existing monocular foundation models to improve the quality of sparse depth inputs, and (2) reformulating training objectives to better capture geometric structure and metric consistency. Specifically, a Poisson-based depth initialization strategy is first introduced to generate a uniform coarse dense depth map from diverse sparse observations, providing a strong structural prior for the network. Regarding the training objective, we replace the conventional depth head with a point map head that regresses per-pixel 3D coordinates in camera space, enabling the model to directly learn the underlying 3D scene structure instead of performing pixel-wise depth map restoration. Moreover, this design eliminates the need for camera intrinsic parameters, allowing LDCM to naturally produce metric-scaled 3D point maps. Extensive experiments demonstrate that LDCM consistently outperforms state-of-the-art methods across multiple benchmarks and varying sparsity levels in both depth completion and point map estimation, showcasing its effectiveness and strong generalization to unseen data distributions.

Related papers

• An End-to-End Room Geometry Constrained Depth Estimation Framework for Indoor Panorama Images [50.84536164535991]
  Existing methods focus on pixel-level accuracy, causing oversmoothed room corners and noise sensitivity.<n>We propose a depth estimation framework based on room geometry constraints.<n>Our framework incorporates two strategies: a room geometry-based background depth resolving strategy and a background-segmentation-guided fusion mechanism.
  arXiv  Detail & Related papers  (2025-10-09T05:52:48Z)
• Propagating Sparse Depth via Depth Foundation Model for Out-of-Distribution Depth Completion [33.854696587141355]
  We propose a novel depth completion framework that leverages depth foundation models to attain remarkable robustness without large-scale training.<n>Specifically, we leverage a depth foundation model to extract environmental cues, including structural and semantic context, from RGB images to guide the propagation of sparse depth information into missing regions.<n>Our framework performs remarkably well in the OOD scenarios and outperforms existing state-of-the-art depth completion methods.
  arXiv  Detail & Related papers  (2025-08-07T02:38:24Z)
• Refinement of Monocular Depth Maps via Multi-View Differentiable Rendering [6.372979654151044]
  Current state-of-the-art monocular depth estimators, trained on extensive datasets, generalize well but lack 3D consistency needed for many applications.<n>In this paper, we combine the strength of those generalizing monocular depth estimation techniques with multi-view data by framing this as an analysis-by-synthesis optimization problem.<n>Our method is able to generate detailed, high-quality, view consistent, accurate depth maps, also in challenging indoor scenarios, and outperforms state-of-the-art multi-view depth reconstruction approaches on such datasets.
  arXiv  Detail & Related papers  (2024-10-04T18:50:28Z)
• GEOcc: Geometrically Enhanced 3D Occupancy Network with Implicit-Explicit Depth Fusion and Contextual Self-Supervision [49.839374549646884]
  This paper presents GEOcc, a Geometric-Enhanced Occupancy network tailored for vision-only surround-view perception.<n>Our approach achieves State-Of-The-Art performance on the Occ3D-nuScenes dataset with the least image resolution needed and the most weightless image backbone.
  arXiv  Detail & Related papers  (2024-05-17T07:31:20Z)
• 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)
• SwinDepth: Unsupervised Depth Estimation using Monocular Sequences via Swin Transformer and Densely Cascaded Network [29.798579906253696]
  It is challenging to acquire dense ground truth depth labels for supervised training, and the unsupervised depth estimation using monocular sequences emerges as a promising alternative. In this paper, we employ a convolution-free Swin Transformer as an image feature extractor so that the network can capture both local geometric features and global semantic features for depth estimation. Also, we propose a Densely Cascaded Multi-scale Network (DCMNet) that connects every feature map directly with another from different scales via a top-down cascade pathway.
  arXiv  Detail & Related papers  (2023-01-17T06:01:46Z)
• MonoJSG: Joint Semantic and Geometric Cost Volume for Monocular 3D Object Detection [10.377424252002792]
  monocular 3D object detection lacks accurate depth recovery ability. Deep neural network (DNN) enables monocular depth-sensing from high-level learned features. We propose a joint semantic and geometric cost volume to model the depth error.
  arXiv  Detail & Related papers  (2022-03-16T11:54:10Z)
• VolumeFusion: Deep Depth Fusion for 3D Scene Reconstruction [71.83308989022635]
  In this paper, we advocate that replicating the traditional two stages framework with deep neural networks improves both the interpretability and the accuracy of the results. Our network operates in two steps: 1) the local computation of the local depth maps with a deep MVS technique, and, 2) the depth maps and images' features fusion to build a single TSDF volume. In order to improve the matching performance between images acquired from very different viewpoints, we introduce a rotation-invariant 3D convolution kernel called PosedConv.
  arXiv  Detail & Related papers  (2021-08-19T11:33:58Z)
• Learning Geometry-Guided Depth via Projective Modeling for Monocular 3D Object Detection [70.71934539556916]
  We learn geometry-guided depth estimation with projective modeling to advance monocular 3D object detection. Specifically, a principled geometry formula with projective modeling of 2D and 3D depth predictions in the monocular 3D object detection network is devised. Our method remarkably improves the detection performance of the state-of-the-art monocular-based method without extra data by 2.80% on the moderate test setting.
  arXiv  Detail & Related papers  (2021-07-29T12:30:39Z)
• Deep Two-View Structure-from-Motion Revisited [83.93809929963969]
  Two-view structure-from-motion (SfM) is the cornerstone of 3D reconstruction and visual SLAM. We propose to revisit the problem of deep two-view SfM by leveraging the well-posedness of the classic pipeline. Our method consists of 1) an optical flow estimation network that predicts dense correspondences between two frames; 2) a normalized pose estimation module that computes relative camera poses from the 2D optical flow correspondences, and 3) a scale-invariant depth estimation network that leverages epipolar geometry to reduce the search space, refine the dense correspondences, and estimate relative depth maps.
  arXiv  Detail & Related papers  (2021-04-01T15:31:20Z)

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.