Dynamic Semantic Occupancy Mapping Using 3D Scene Flow and Closed-Form Bayesian Inference
This paper reports on a dynamic semantic mapping framework that incorporates 3D scene flow measurements into a closed-form Bayesian inference model. Existence of dynamic objects in the environment can cause artifacts and traces in current mapping algorithms, leading to an inconsistent map posterior....
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2022-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9882042/ |
| _version_ | 1828110740078198784 |
|---|---|
| author | Aishwarya Unnikrishnan Joey Wilson Lu Gan Andrew Capodieci Paramsothy Jayakumar Kira Barton Maani Ghaffari |
| author_facet | Aishwarya Unnikrishnan Joey Wilson Lu Gan Andrew Capodieci Paramsothy Jayakumar Kira Barton Maani Ghaffari |
| author_sort | Aishwarya Unnikrishnan |
| collection | DOAJ |
| description | This paper reports on a dynamic semantic mapping framework that incorporates 3D scene flow measurements into a closed-form Bayesian inference model. Existence of dynamic objects in the environment can cause artifacts and traces in current mapping algorithms, leading to an inconsistent map posterior. We leverage state-of-the-art semantic segmentation and 3D flow estimation using deep learning to provide measurements for map inference. We develop a Bayesian model that propagates the scene with flow and infers a 3D continuous (i.e., can be queried at arbitrary resolution) semantic occupancy map outperforming its static counterpart. Extensive experiments using publicly available data sets show that the proposed framework improves over its predecessors and input measurements from deep neural networks consistently. |
| first_indexed | 2024-04-11T11:24:01Z |
| format | Article |
| id | doaj.art-a86f7ea8fee14241a350ed9a86d301fd |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-04-11T11:24:01Z |
| publishDate | 2022-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-a86f7ea8fee14241a350ed9a86d301fd2022-12-22T04:26:27ZengIEEEIEEE Access2169-35362022-01-0110979549797010.1109/ACCESS.2022.32053299882042Dynamic Semantic Occupancy Mapping Using 3D Scene Flow and Closed-Form Bayesian InferenceAishwarya Unnikrishnan0Joey Wilson1https://orcid.org/0000-0002-9305-9647Lu Gan2https://orcid.org/0000-0003-0911-8032Andrew Capodieci3Paramsothy Jayakumar4Kira Barton5https://orcid.org/0000-0003-1047-8078Maani Ghaffari6https://orcid.org/0000-0002-4734-4295Department of Robotics, University of Michigan, Ann Arbor, MI, USADepartment of Robotics, University of Michigan, Ann Arbor, MI, USADepartment of Robotics, University of Michigan, Ann Arbor, MI, USANeya Systems Division, Applied Research Associates, Pittsburgh, PA, USAU.S. Army CCDC Ground Vehicle Systems Center, Warren, MI, USADepartment of Robotics, University of Michigan, Ann Arbor, MI, USADepartment of Robotics, University of Michigan, Ann Arbor, MI, USAThis paper reports on a dynamic semantic mapping framework that incorporates 3D scene flow measurements into a closed-form Bayesian inference model. Existence of dynamic objects in the environment can cause artifacts and traces in current mapping algorithms, leading to an inconsistent map posterior. We leverage state-of-the-art semantic segmentation and 3D flow estimation using deep learning to provide measurements for map inference. We develop a Bayesian model that propagates the scene with flow and infers a 3D continuous (i.e., can be queried at arbitrary resolution) semantic occupancy map outperforming its static counterpart. Extensive experiments using publicly available data sets show that the proposed framework improves over its predecessors and input measurements from deep neural networks consistently.https://ieeexplore.ieee.org/document/9882042/Bayesian inferencecomputer visionmappingsemantic scene understanding |
| spellingShingle | Aishwarya Unnikrishnan Joey Wilson Lu Gan Andrew Capodieci Paramsothy Jayakumar Kira Barton Maani Ghaffari Dynamic Semantic Occupancy Mapping Using 3D Scene Flow and Closed-Form Bayesian Inference IEEE Access Bayesian inference computer vision mapping semantic scene understanding |
| title | Dynamic Semantic Occupancy Mapping Using 3D Scene Flow and Closed-Form Bayesian Inference |
| title_full | Dynamic Semantic Occupancy Mapping Using 3D Scene Flow and Closed-Form Bayesian Inference |
| title_fullStr | Dynamic Semantic Occupancy Mapping Using 3D Scene Flow and Closed-Form Bayesian Inference |
| title_full_unstemmed | Dynamic Semantic Occupancy Mapping Using 3D Scene Flow and Closed-Form Bayesian Inference |
| title_short | Dynamic Semantic Occupancy Mapping Using 3D Scene Flow and Closed-Form Bayesian Inference |
| title_sort | dynamic semantic occupancy mapping using 3d scene flow and closed form bayesian inference |
| topic | Bayesian inference computer vision mapping semantic scene understanding |
| url | https://ieeexplore.ieee.org/document/9882042/ |
| work_keys_str_mv | AT aishwaryaunnikrishnan dynamicsemanticoccupancymappingusing3dsceneflowandclosedformbayesianinference AT joeywilson dynamicsemanticoccupancymappingusing3dsceneflowandclosedformbayesianinference AT lugan dynamicsemanticoccupancymappingusing3dsceneflowandclosedformbayesianinference AT andrewcapodieci dynamicsemanticoccupancymappingusing3dsceneflowandclosedformbayesianinference AT paramsothyjayakumar dynamicsemanticoccupancymappingusing3dsceneflowandclosedformbayesianinference AT kirabarton dynamicsemanticoccupancymappingusing3dsceneflowandclosedformbayesianinference AT maanighaffari dynamicsemanticoccupancymappingusing3dsceneflowandclosedformbayesianinference |