Incremental Joint Learning of Depth, Pose and Implicit Scene Representation on Monocular Camera in Large-scale Scenes

• URL: http://arxiv.org/abs/2404.06050v1
• Date: Tue, 9 Apr 2024 06:27:35 GMT
• Title: Incremental Joint Learning of Depth, Pose and Implicit Scene Representation on Monocular Camera in Large-scale Scenes
• Authors: Tianchen Deng, Nailin Wang, Chongdi Wang, Shenghai Yuan, Jingchuan Wang, Danwei Wang, Weidong Chen,
• Abstract summary: We propose an incremental joint learning framework, which can achieve accurate depth, pose estimation, and large-scale scene reconstruction. A vision transformer-based network is adopted as the backbone to enhance performance in scale information estimation. In terms of implicit scene representation, we propose an incremental scene representation method to construct the entire large-scale scene as multiple local radiance fields.
• Score: 17.055674468733244
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Dense scene reconstruction for photo-realistic view synthesis has various applications, such as VR/AR, autonomous vehicles. However, most existing methods have difficulties in large-scale scenes due to three core challenges: \textit{(a) inaccurate depth input.} Accurate depth input is impossible to get in real-world large-scale scenes. \textit{(b) inaccurate pose estimation.} Most existing approaches rely on accurate pre-estimated camera poses. \textit{(c) insufficient scene representation capability.} A single global radiance field lacks the capacity to effectively scale to large-scale scenes. To this end, we propose an incremental joint learning framework, which can achieve accurate depth, pose estimation, and large-scale scene reconstruction. A vision transformer-based network is adopted as the backbone to enhance performance in scale information estimation. For pose estimation, a feature-metric bundle adjustment (FBA) method is designed for accurate and robust camera tracking in large-scale scenes. In terms of implicit scene representation, we propose an incremental scene representation method to construct the entire large-scale scene as multiple local radiance fields to enhance the scalability of 3D scene representation. Extended experiments have been conducted to demonstrate the effectiveness and accuracy of our method in depth estimation, pose estimation, and large-scale scene reconstruction.

Related papers

• ScaleDepth: Decomposing Metric Depth Estimation into Scale Prediction and Relative Depth Estimation [62.600382533322325]
  We propose a novel monocular depth estimation method called ScaleDepth. Our method decomposes metric depth into scene scale and relative depth, and predicts them through a semantic-aware scale prediction module. Our method achieves metric depth estimation for both indoor and outdoor scenes in a unified framework.
  arXiv  Detail & Related papers  (2024-07-11T05:11:56Z)
• GLACE: Global Local Accelerated Coordinate Encoding [66.87005863868181]
  Scene coordinate regression methods are effective in small-scale scenes but face significant challenges in large-scale scenes. We propose GLACE, which integrates pre-trained global and local encodings and enables SCR to scale to large scenes with only a single small-sized network. Our method achieves state-of-the-art results on large-scale scenes with a low-map-size model.
  arXiv  Detail & Related papers  (2024-06-06T17:59:50Z)
• 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)
• FlowCam: Training Generalizable 3D Radiance Fields without Camera Poses via Pixel-Aligned Scene Flow [26.528667940013598]
  Reconstruction of 3D neural fields from posed images has emerged as a promising method for self-supervised representation learning. Key challenge preventing the deployment of these 3D scene learners on large-scale video data is their dependence on precise camera poses from structure-from-motion. We propose a method that jointly reconstructs camera poses and 3D neural scene representations online and in a single forward pass.
  arXiv  Detail & Related papers  (2023-05-31T20:58:46Z)
• Progressively Optimized Local Radiance Fields for Robust View Synthesis [76.55036080270347]
  We present an algorithm for reconstructing the radiance field of a large-scale scene from a single casually captured video. For handling unknown poses, we jointly estimate the camera poses with radiance field in a progressive manner. For handling large unbounded scenes, we dynamically allocate new local radiance fields trained with frames within a temporal window.
  arXiv  Detail & Related papers  (2023-03-24T04:03:55Z)
• RUST: Latent Neural Scene Representations from Unposed Imagery [21.433079925439234]
  Inferring structure of 3D scenes from 2D observations is a fundamental challenge in computer vision. Recent popularized approaches based on neural scene representations have achieved tremendous impact. RUST (Really Unposed Scene representation Transformer) is a pose-free approach to novel view trained on RGB images alone.
  arXiv  Detail & Related papers  (2022-11-25T18:59:10Z)
• Towards Accurate Reconstruction of 3D Scene Shape from A Single Monocular Image [91.71077190961688]
  We propose a two-stage framework that first predicts depth up to an unknown scale and shift from a single monocular image. We then exploits 3D point cloud data to predict the depth shift and the camera's focal length that allow us to recover 3D scene shapes. We test our depth model on nine unseen datasets and achieve state-of-the-art performance on zero-shot evaluation.
  arXiv  Detail & Related papers  (2022-08-28T16:20:14Z)
• SCONE: Surface Coverage Optimization in Unknown Environments by Volumetric Integration [23.95135709027516]
  Next Best View computation (NBV) is a long-standing problem in robotics. We show that we can maximize surface metrics by Monte Carlo integration over a volumetric representation. It takes as input an arbitrarily large point cloud gathered by a depth sensor like Lidar systems as well as camera poses to predict NBV.
  arXiv  Detail & Related papers  (2022-08-22T17:04:14Z)
• Towards 3D Scene Reconstruction from Locally Scale-Aligned Monocular Video Depth [90.33296913575818]
  In some video-based scenarios such as video depth estimation and 3D scene reconstruction from a video, the unknown scale and shift residing in per-frame prediction may cause the depth inconsistency. We propose a locally weighted linear regression method to recover the scale and shift with very sparse anchor points. Our method can boost the performance of existing state-of-the-art approaches by 50% at most over several zero-shot benchmarks.
  arXiv  Detail & Related papers  (2022-02-03T08:52:54Z)

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.