PanoSCU: A Simulation-Based Dataset for Panoramic Indoor Scene Understanding

Panoramic images offer a comprehensive spatial view that is crucial for indoor robotics tasks such as visual room rearrangement, where an agent must restore objects to their original positions or states. Unlike existing 2D scene change understanding datasets, which rely on single-view images, panora...

Full description

Saved in:
Bibliographic Details
Main Authors: Mariia Khan, Yue Qiu, Yuren Cong, Jumana Abu-Khalaf, David Suter, Bodo Rosenhahn
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10965672/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849712691534364672
author Mariia Khan
Yue Qiu
Yuren Cong
Jumana Abu-Khalaf
David Suter
Bodo Rosenhahn
author_facet Mariia Khan
Yue Qiu
Yuren Cong
Jumana Abu-Khalaf
David Suter
Bodo Rosenhahn
author_sort Mariia Khan
collection DOAJ
description Panoramic images offer a comprehensive spatial view that is crucial for indoor robotics tasks such as visual room rearrangement, where an agent must restore objects to their original positions or states. Unlike existing 2D scene change understanding datasets, which rely on single-view images, panoramic views capture richer spatial context, object relationships, and occlusions—making them better suited for embodied artificial intelligence (AI) applications. To address this, we introduce Panoramic Scene Change Understanding (PanoSCU), a dataset specifically designed to enhance the visual object rearrangement task. Our dataset comprises 5,300 panoramas generated in an embodied simulator, encompassing 48 common indoor object classes. PanoSCU supports eight research tasks: single-view and panoramic detection, single-view and panoramic segmentation, single-view and panoramic change understanding, embodied object tracking, and change reversal. We also present PanoStitch, a training-free method for automatic panoramic data collection within embodied environments. We evaluate state-of-the-art methods on panoramic segmentation and change understanding tasks. There is a gap in existing methods, as they are not designed for panoramic inputs and struggle with varying ratios and sizes, resulting from the unique challenges of visual object rearrangement. Our findings reveal these limitations and underscore PanoSCU’s potential to drive progress in developing models capable of robust panoramic reasoning and fine-grained scene change understanding.
format Article
id doaj-art-19825937e4ea45b9bb18fb83e4c971d5
institution DOAJ
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-19825937e4ea45b9bb18fb83e4c971d52025-08-20T03:14:12ZengIEEEIEEE Access2169-35362025-01-0113724567247610.1109/ACCESS.2025.356105510965672PanoSCU: A Simulation-Based Dataset for Panoramic Indoor Scene UnderstandingMariia Khan0https://orcid.org/0000-0001-6662-4607Yue Qiu1Yuren Cong2Jumana Abu-Khalaf3https://orcid.org/0000-0002-6651-2880David Suter4https://orcid.org/0000-0001-6306-3023Bodo Rosenhahn5https://orcid.org/0000-0003-3861-1424School of Science, Centre for AI and Machine Learning, Edith Cowan University, Joondalup, WA, AustraliaArtificial Intelligence Research Center, National Institute of Advanced Industrial Science and Technology, Chiyoda, JapanInstitute for Information Processing, Leibniz University Hannover, Hannover, GermanySchool of Science, Centre for AI and Machine Learning, Edith Cowan University, Joondalup, WA, AustraliaSchool of Science, Centre for AI and Machine Learning, Edith Cowan University, Joondalup, WA, AustraliaInstitute for Information Processing, Leibniz University Hannover, Hannover, GermanyPanoramic images offer a comprehensive spatial view that is crucial for indoor robotics tasks such as visual room rearrangement, where an agent must restore objects to their original positions or states. Unlike existing 2D scene change understanding datasets, which rely on single-view images, panoramic views capture richer spatial context, object relationships, and occlusions—making them better suited for embodied artificial intelligence (AI) applications. To address this, we introduce Panoramic Scene Change Understanding (PanoSCU), a dataset specifically designed to enhance the visual object rearrangement task. Our dataset comprises 5,300 panoramas generated in an embodied simulator, encompassing 48 common indoor object classes. PanoSCU supports eight research tasks: single-view and panoramic detection, single-view and panoramic segmentation, single-view and panoramic change understanding, embodied object tracking, and change reversal. We also present PanoStitch, a training-free method for automatic panoramic data collection within embodied environments. We evaluate state-of-the-art methods on panoramic segmentation and change understanding tasks. There is a gap in existing methods, as they are not designed for panoramic inputs and struggle with varying ratios and sizes, resulting from the unique challenges of visual object rearrangement. Our findings reveal these limitations and underscore PanoSCU’s potential to drive progress in developing models capable of robust panoramic reasoning and fine-grained scene change understanding.https://ieeexplore.ieee.org/document/10965672/Object segmentationchange detection algorithmsembodied artificial intelligenceimage stitching
spellingShingle Mariia Khan
Yue Qiu
Yuren Cong
Jumana Abu-Khalaf
David Suter
Bodo Rosenhahn
PanoSCU: A Simulation-Based Dataset for Panoramic Indoor Scene Understanding
IEEE Access
Object segmentation
change detection algorithms
embodied artificial intelligence
image stitching
title PanoSCU: A Simulation-Based Dataset for Panoramic Indoor Scene Understanding
title_full PanoSCU: A Simulation-Based Dataset for Panoramic Indoor Scene Understanding
title_fullStr PanoSCU: A Simulation-Based Dataset for Panoramic Indoor Scene Understanding
title_full_unstemmed PanoSCU: A Simulation-Based Dataset for Panoramic Indoor Scene Understanding
title_short PanoSCU: A Simulation-Based Dataset for Panoramic Indoor Scene Understanding
title_sort panoscu a simulation based dataset for panoramic indoor scene understanding
topic Object segmentation
change detection algorithms
embodied artificial intelligence
image stitching
url https://ieeexplore.ieee.org/document/10965672/
work_keys_str_mv AT mariiakhan panoscuasimulationbaseddatasetforpanoramicindoorsceneunderstanding
AT yueqiu panoscuasimulationbaseddatasetforpanoramicindoorsceneunderstanding
AT yurencong panoscuasimulationbaseddatasetforpanoramicindoorsceneunderstanding
AT jumanaabukhalaf panoscuasimulationbaseddatasetforpanoramicindoorsceneunderstanding
AT davidsuter panoscuasimulationbaseddatasetforpanoramicindoorsceneunderstanding
AT bodorosenhahn panoscuasimulationbaseddatasetforpanoramicindoorsceneunderstanding