Flow-NeRF: Joint Learning of Geometry, Poses, and Dense Flow within Unified Neural Representations Article Swipe

PDF

YOU? · · 2025 · Open Access · · DOI: https://doi.org/10.48550/arxiv.2503.10464

Learning accurate scene reconstruction without pose priors in neural radiance fields is challenging due to inherent geometric ambiguity. Recent development either relies on correspondence priors for regularization or uses off-the-shelf flow estimators to derive analytical poses. However, the potential for jointly learning scene geometry, camera poses, and dense flow within a unified neural representation remains largely unexplored. In this paper, we present Flow-NeRF, a unified framework that simultaneously optimizes scene geometry, camera poses, and dense optical flow all on-the-fly. To enable the learning of dense flow within the neural radiance field, we design and build a bijective mapping for flow estimation, conditioned on pose. To make the scene reconstruction benefit from the flow estimation, we develop an effective feature enhancement mechanism to pass canonical space features to world space representations, significantly enhancing scene geometry. We validate our model across four important tasks, i.e., novel view synthesis, depth estimation, camera pose prediction, and dense optical flow estimation, using several datasets. Our approach surpasses previous methods in almost all metrics for novel-view view synthesis and depth estimation and yields both qualitatively sound and quantitatively accurate novel-view flow. Our project page is https://zhengxunzhi.github.io/flownerf/.

Related Topics

Maya Joint

Learning Curve

Experiential Learning

Practice (Learning Method)

Learning Environment

Machine Learning

Deep Learning

Reinforcement Learning

Joint Strike Missile

Learning Standards

Joint Terminal Attack Controller

Agm-154 Joint Standoff Weapon

Concepts

No concepts available.

Metadata

Type: preprint
Language: en
Landing Page: http://arxiv.org/abs/2503.10464
PDF: https://arxiv.org/pdf/2503.10464
OA Status: green
OpenAlex ID: https://openalex.org/W4416039555

All OpenAlex metadata

Raw OpenAlex JSON

OpenAlex ID: https://openalex.org/W4416039555

Canonical identifier for this work in OpenAlex
DOI: https://doi.org/10.48550/arxiv.2503.10464

Digital Object Identifier
Title: Flow-NeRF: Joint Learning of Geometry, Poses, and Dense Flow within Unified Neural Representations

Work title
Type: preprint

OpenAlex work type
Language: en

Primary language
Publication year: 2025

Year of publication
Publication date: 2025-03-13

Full publication date if available
Authors: Xiaoming Zheng, Dan Xu

List of authors in order
Landing page: https://arxiv.org/abs/2503.10464

Publisher landing page
PDF URL: https://arxiv.org/pdf/2503.10464

Direct link to full text PDF
Open access: Yes

Whether a free full text is available
OA status: green

Open access status per OpenAlex
OA URL: https://arxiv.org/pdf/2503.10464

Direct OA link when available
Cited by: 0

Total citation count in OpenAlex

Full payload

id	https://openalex.org/W4416039555
doi	https://doi.org/10.48550/arxiv.2503.10464
ids.doi	https://doi.org/10.48550/arxiv.2503.10464
ids.openalex	https://openalex.org/W4416039555
fwci
type	preprint
title	Flow-NeRF: Joint Learning of Geometry, Poses, and Dense Flow within Unified Neural Representations
biblio.issue
biblio.volume
biblio.last_page
biblio.first_page
is_xpac	False
apc_list
apc_paid
language	en
locations[0].id	pmh:oai:arXiv.org:2503.10464
locations[0].is_oa	True
locations[0].source.id	https://openalex.org/S4306400194
locations[0].source.issn
locations[0].source.type	repository
locations[0].source.is_oa	True
locations[0].source.issn_l
locations[0].source.is_core	False
locations[0].source.is_in_doaj	False
locations[0].source.display_name	arXiv (Cornell University)
locations[0].source.host_organization	https://openalex.org/I205783295
locations[0].source.host_organization_name	Cornell University
locations[0].source.host_organization_lineage	https://openalex.org/I205783295
locations[0].license
locations[0].pdf_url	https://arxiv.org/pdf/2503.10464
locations[0].version	submittedVersion
locations[0].raw_type	text
locations[0].license_id
locations[0].is_accepted	False
locations[0].is_published	False
locations[0].raw_source_name
locations[0].landing_page_url	http://arxiv.org/abs/2503.10464
locations[1].id	doi:10.48550/arxiv.2503.10464
locations[1].is_oa	True
locations[1].source.id	https://openalex.org/S4306400194
locations[1].source.issn
locations[1].source.type	repository
locations[1].source.is_oa	True
locations[1].source.issn_l
locations[1].source.is_core	False
locations[1].source.is_in_doaj	False
locations[1].source.display_name	arXiv (Cornell University)
locations[1].source.host_organization	https://openalex.org/I205783295
locations[1].source.host_organization_name	Cornell University
locations[1].source.host_organization_lineage	https://openalex.org/I205783295
locations[1].license
locations[1].pdf_url
locations[1].version
locations[1].raw_type	article
locations[1].license_id
locations[1].is_accepted	False
locations[1].is_published
locations[1].raw_source_name
locations[1].landing_page_url	https://doi.org/10.48550/arxiv.2503.10464
indexed_in	arxiv, datacite
authorships[0].author.id	https://openalex.org/A5081509849
authorships[0].author.orcid	https://orcid.org/0000-0003-0918-2624
authorships[0].author.display_name	Xiaoming Zheng
authorships[0].author_position	first
authorships[0].raw_author_name	Zheng, Xunzhi
authorships[0].is_corresponding	False
authorships[1].author.id	https://openalex.org/A5100341938
authorships[1].author.orcid	https://orcid.org/0000-0003-0136-9603
authorships[1].author.display_name	Dan Xu
authorships[1].author_position	last
authorships[1].raw_author_name	Xu, Dan
authorships[1].is_corresponding	False
has_content.pdf	False
has_content.grobid_xml	False
is_paratext	False
open_access.is_oa	True
open_access.oa_url	https://arxiv.org/pdf/2503.10464
open_access.oa_status	green
open_access.any_repository_has_fulltext	False
created_date	2025-10-10T00:00:00
display_name	Flow-NeRF: Joint Learning of Geometry, Poses, and Dense Flow within Unified Neural Representations
has_fulltext	False
is_retracted	False
updated_date	2025-11-09T23:09:16.995542
primary_topic
cited_by_count	0
locations_count	2
best_oa_location.id	pmh:oai:arXiv.org:2503.10464
best_oa_location.is_oa	True
best_oa_location.source.id	https://openalex.org/S4306400194
best_oa_location.source.issn
best_oa_location.source.type	repository
best_oa_location.source.is_oa	True
best_oa_location.source.issn_l
best_oa_location.source.is_core	False
best_oa_location.source.is_in_doaj	False
best_oa_location.source.display_name	arXiv (Cornell University)
best_oa_location.source.host_organization	https://openalex.org/I205783295
best_oa_location.source.host_organization_name	Cornell University
best_oa_location.source.host_organization_lineage	https://openalex.org/I205783295
best_oa_location.license
best_oa_location.pdf_url	https://arxiv.org/pdf/2503.10464
best_oa_location.version	submittedVersion
best_oa_location.raw_type	text
best_oa_location.license_id
best_oa_location.is_accepted	False
best_oa_location.is_published	False
best_oa_location.raw_source_name
best_oa_location.landing_page_url	http://arxiv.org/abs/2503.10464
primary_location.id	pmh:oai:arXiv.org:2503.10464
primary_location.is_oa	True
primary_location.source.id	https://openalex.org/S4306400194
primary_location.source.issn
primary_location.source.type	repository
primary_location.source.is_oa	True
primary_location.source.issn_l
primary_location.source.is_core	False
primary_location.source.is_in_doaj	False
primary_location.source.display_name	arXiv (Cornell University)
primary_location.source.host_organization	https://openalex.org/I205783295
primary_location.source.host_organization_name	Cornell University
primary_location.source.host_organization_lineage	https://openalex.org/I205783295
primary_location.license
primary_location.pdf_url	https://arxiv.org/pdf/2503.10464
primary_location.version	submittedVersion
primary_location.raw_type	text
primary_location.license_id
primary_location.is_accepted	False
primary_location.is_published	False
primary_location.raw_source_name
primary_location.landing_page_url	http://arxiv.org/abs/2503.10464
publication_date	2025-03-13
publication_year	2025
referenced_works_count	0
abstract_inverted_index.a	50, 63, 95
abstract_inverted_index.In	57
abstract_inverted_index.To	79, 104
abstract_inverted_index.We	134
abstract_inverted_index.an	116
abstract_inverted_index.in	7, 164
abstract_inverted_index.is	11, 188
abstract_inverted_index.of	83
abstract_inverted_index.on	22, 102
abstract_inverted_index.or	27
abstract_inverted_index.to	14, 32, 121, 126
abstract_inverted_index.we	60, 91, 114
abstract_inverted_index.Our	159, 185
abstract_inverted_index.all	77, 166
abstract_inverted_index.and	46, 73, 93, 151, 172, 175, 180
abstract_inverted_index.due	13
abstract_inverted_index.for	25, 39, 98, 168
abstract_inverted_index.our	136
abstract_inverted_index.the	37, 81, 87, 106, 111
abstract_inverted_index.both	177
abstract_inverted_index.flow	30, 48, 76, 85, 99, 112, 154
abstract_inverted_index.four	139
abstract_inverted_index.from	110
abstract_inverted_index.make	105
abstract_inverted_index.page	187
abstract_inverted_index.pass	122
abstract_inverted_index.pose	5, 149
abstract_inverted_index.that	66
abstract_inverted_index.this	58
abstract_inverted_index.uses	28
abstract_inverted_index.view	144, 170
abstract_inverted_index.build	94
abstract_inverted_index.dense	47, 74, 84, 152
abstract_inverted_index.depth	146, 173
abstract_inverted_index.flow.	184
abstract_inverted_index.i.e.,	142
abstract_inverted_index.model	137
abstract_inverted_index.novel	143
abstract_inverted_index.pose.	103
abstract_inverted_index.scene	2, 42, 69, 107, 132
abstract_inverted_index.sound	179
abstract_inverted_index.space	124, 128
abstract_inverted_index.using	156
abstract_inverted_index.world	127
abstract_inverted_index.Recent	18
abstract_inverted_index.across	138
abstract_inverted_index.almost	165
abstract_inverted_index.camera	44, 71, 148
abstract_inverted_index.derive	33
abstract_inverted_index.design	92
abstract_inverted_index.either	20
abstract_inverted_index.enable	80
abstract_inverted_index.field,	90
abstract_inverted_index.fields	10
abstract_inverted_index.neural	8, 52, 88
abstract_inverted_index.paper,	59
abstract_inverted_index.poses,	45, 72
abstract_inverted_index.poses.	35
abstract_inverted_index.priors	6, 24
abstract_inverted_index.relies	21
abstract_inverted_index.tasks,	141
abstract_inverted_index.within	49, 86
abstract_inverted_index.yields	176
abstract_inverted_index.benefit	109
abstract_inverted_index.develop	115
abstract_inverted_index.feature	118
abstract_inverted_index.jointly	40
abstract_inverted_index.largely	55
abstract_inverted_index.mapping	97
abstract_inverted_index.methods	163
abstract_inverted_index.metrics	167
abstract_inverted_index.optical	75, 153
abstract_inverted_index.present	61
abstract_inverted_index.project	186
abstract_inverted_index.remains	54
abstract_inverted_index.several	157
abstract_inverted_index.unified	51, 64
abstract_inverted_index.without	4
abstract_inverted_index.However,	36
abstract_inverted_index.Learning	0
abstract_inverted_index.accurate	1, 182
abstract_inverted_index.approach	160
abstract_inverted_index.features	125
abstract_inverted_index.inherent	15
abstract_inverted_index.learning	41, 82
abstract_inverted_index.previous	162
abstract_inverted_index.radiance	9, 89
abstract_inverted_index.validate	135
abstract_inverted_index.bijective	96
abstract_inverted_index.canonical	123
abstract_inverted_index.datasets.	158
abstract_inverted_index.effective	117
abstract_inverted_index.enhancing	131
abstract_inverted_index.framework	65
abstract_inverted_index.geometric	16
abstract_inverted_index.geometry,	43, 70
abstract_inverted_index.geometry.	133
abstract_inverted_index.important	140
abstract_inverted_index.mechanism	120
abstract_inverted_index.optimizes	68
abstract_inverted_index.potential	38
abstract_inverted_index.surpasses	161
abstract_inverted_index.synthesis	171
abstract_inverted_index.Flow-NeRF,	62
abstract_inverted_index.ambiguity.	17
abstract_inverted_index.analytical	34
abstract_inverted_index.estimation	174
abstract_inverted_index.estimators	31
abstract_inverted_index.novel-view	169, 183
abstract_inverted_index.synthesis,	145
abstract_inverted_index.challenging	12
abstract_inverted_index.conditioned	101
abstract_inverted_index.development	19
abstract_inverted_index.enhancement	119
abstract_inverted_index.estimation,	100, 113, 147, 155
abstract_inverted_index.on-the-fly.	78
abstract_inverted_index.prediction,	150
abstract_inverted_index.unexplored.	56
abstract_inverted_index.off-the-shelf	29
abstract_inverted_index.qualitatively	178
abstract_inverted_index.significantly	130
abstract_inverted_index.correspondence	23
abstract_inverted_index.quantitatively	181
abstract_inverted_index.reconstruction	3, 108
abstract_inverted_index.regularization	26
abstract_inverted_index.representation	53
abstract_inverted_index.simultaneously	67
abstract_inverted_index.representations,	129
abstract_inverted_index.https://zhengxunzhi.github.io/flownerf/.	189
cited_by_percentile_year
countries_distinct_count	0
institutions_distinct_count	2
citation_normalized_percentile