OmniOcc: Cylindrical Voxel-Based Semantic Occupancy Prediction for Omnidirectional Vision Systems

Accurate 3D perception is essential for autonomous driving. Traditional methods often struggle with geometric ambiguity due to a lack of geometric prior. To address these challenges, we use omnidirectional depth estimation to introduce geometric prior. Based on the depth information, we propose a Sk...

Full description

Saved in:
Bibliographic Details
Main Authors: Chaofan Wu, Jiaheng Li, Jinghao Cao, Ming Li, Sidan Du, Yang Li
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11113275/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849390432508706816
author Chaofan Wu
Jiaheng Li
Jinghao Cao
Ming Li
Sidan Du
Yang Li
author_facet Chaofan Wu
Jiaheng Li
Jinghao Cao
Ming Li
Sidan Du
Yang Li
author_sort Chaofan Wu
collection DOAJ
description Accurate 3D perception is essential for autonomous driving. Traditional methods often struggle with geometric ambiguity due to a lack of geometric prior. To address these challenges, we use omnidirectional depth estimation to introduce geometric prior. Based on the depth information, we propose a Sketch-Coloring framework OmniOcc. Additionally, our approach introduces a cylindrical voxel representation based on polar coordinate to better align with the radial nature of panoramic camera views. To address the lack of fisheye camera dataset in autonomous driving tasks, we also build a virtual scene dataset with six fisheye cameras. Experimental results demonstrate that our Sketch-Coloring network significantly enhances 3D perception performance.
format Article
id doaj-art-79e80d2dff6344bfb13b3628ae9d9329
institution Kabale University
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-79e80d2dff6344bfb13b3628ae9d93292025-08-20T03:41:40ZengIEEEIEEE Access2169-35362025-01-011313994413995210.1109/ACCESS.2025.359589811113275OmniOcc: Cylindrical Voxel-Based Semantic Occupancy Prediction for Omnidirectional Vision SystemsChaofan Wu0https://orcid.org/0009-0004-7215-329XJiaheng Li1Jinghao Cao2https://orcid.org/0000-0002-9408-4553Ming Li3https://orcid.org/0000-0002-1341-5585Sidan Du4https://orcid.org/0000-0002-9966-3765Yang Li5https://orcid.org/0000-0001-6769-4076School of Electronic Science and Engineering, Nanjing University, Nanjing, ChinaSchool of Electronic Science and Engineering, Nanjing University, Nanjing, ChinaSchool of Electronic Science and Engineering, Nanjing University, Nanjing, ChinaSchool of Electronic Science and Engineering, Nanjing University, Nanjing, ChinaSchool of Electronic Science and Engineering, Nanjing University, Nanjing, ChinaSchool of Electronic Science and Engineering, Nanjing University, Nanjing, ChinaAccurate 3D perception is essential for autonomous driving. Traditional methods often struggle with geometric ambiguity due to a lack of geometric prior. To address these challenges, we use omnidirectional depth estimation to introduce geometric prior. Based on the depth information, we propose a Sketch-Coloring framework OmniOcc. Additionally, our approach introduces a cylindrical voxel representation based on polar coordinate to better align with the radial nature of panoramic camera views. To address the lack of fisheye camera dataset in autonomous driving tasks, we also build a virtual scene dataset with six fisheye cameras. Experimental results demonstrate that our Sketch-Coloring network significantly enhances 3D perception performance.https://ieeexplore.ieee.org/document/11113275/Omnidirectional depthcylindrical voxel representationfisheye cameraoccupancy
spellingShingle Chaofan Wu
Jiaheng Li
Jinghao Cao
Ming Li
Sidan Du
Yang Li
OmniOcc: Cylindrical Voxel-Based Semantic Occupancy Prediction for Omnidirectional Vision Systems
IEEE Access
Omnidirectional depth
cylindrical voxel representation
fisheye camera
occupancy
title OmniOcc: Cylindrical Voxel-Based Semantic Occupancy Prediction for Omnidirectional Vision Systems
title_full OmniOcc: Cylindrical Voxel-Based Semantic Occupancy Prediction for Omnidirectional Vision Systems
title_fullStr OmniOcc: Cylindrical Voxel-Based Semantic Occupancy Prediction for Omnidirectional Vision Systems
title_full_unstemmed OmniOcc: Cylindrical Voxel-Based Semantic Occupancy Prediction for Omnidirectional Vision Systems
title_short OmniOcc: Cylindrical Voxel-Based Semantic Occupancy Prediction for Omnidirectional Vision Systems
title_sort omniocc cylindrical voxel based semantic occupancy prediction for omnidirectional vision systems
topic Omnidirectional depth
cylindrical voxel representation
fisheye camera
occupancy
url https://ieeexplore.ieee.org/document/11113275/
work_keys_str_mv AT chaofanwu omniocccylindricalvoxelbasedsemanticoccupancypredictionforomnidirectionalvisionsystems
AT jiahengli omniocccylindricalvoxelbasedsemanticoccupancypredictionforomnidirectionalvisionsystems
AT jinghaocao omniocccylindricalvoxelbasedsemanticoccupancypredictionforomnidirectionalvisionsystems
AT mingli omniocccylindricalvoxelbasedsemanticoccupancypredictionforomnidirectionalvisionsystems
AT sidandu omniocccylindricalvoxelbasedsemanticoccupancypredictionforomnidirectionalvisionsystems
AT yangli omniocccylindricalvoxelbasedsemanticoccupancypredictionforomnidirectionalvisionsystems