Universal Multi-Domain Translation via Diffusion Routers Article Swipe

PDF

Duc Kieu , Kien Do , Tuan Hoang , Thao Minh Le , Tung Kieu , Dang H. Nguyen , Thin Nguyen ·

YOU? · · 2025 · Open Access · · DOI: https://doi.org/10.48550/arxiv.2510.03252

Multi-domain translation (MDT) aims to learn translations between multiple domains, yet existing approaches either require fully aligned tuples or can only handle domain pairs seen in training, limiting their practicality and excluding many cross-domain mappings. We introduce universal MDT (UMDT), a generalization of MDT that seeks to translate between any pair of $K$ domains using only $K-1$ paired datasets with a central domain. To tackle this problem, we propose Diffusion Router (DR), a unified diffusion-based framework that models all central$\leftrightarrow$non-central translations with a single noise predictor conditioned on the source and target domain labels. DR enables indirect non-central translations by routing through the central domain. We further introduce a novel scalable learning strategy with a variational-bound objective and an efficient Tweedie refinement procedure to support direct non-central mappings. Through evaluation on three large-scale UMDT benchmarks, DR achieves state-of-the-art results for both indirect and direct translations, while lowering sampling cost and unlocking novel tasks such as sketch$\leftrightarrow$segmentation. These results establish DR as a scalable and versatile framework for universal translation across multiple domains.

Related Topics

Machine Translation

Universal Store

Lost In Translation (Film)

Universal Epic Universe

Universal Music Group

Universal Declaration Of Human Rights

Church Universal And Triumphant

Last Universal Common Ancestor

Universal Monsters

Legal Translation

Universal Basic Income

Universal Studios Florida

Concepts

No concepts available.

Metadata

Type: preprint
Language: en
Landing Page: http://arxiv.org/abs/2510.03252
PDF: https://arxiv.org/pdf/2510.03252
OA Status: green
OpenAlex ID: https://openalex.org/W4414968558

All OpenAlex metadata

Raw OpenAlex JSON

OpenAlex ID: https://openalex.org/W4414968558

Canonical identifier for this work in OpenAlex
DOI: https://doi.org/10.48550/arxiv.2510.03252

Digital Object Identifier
Title: Universal Multi-Domain Translation via Diffusion Routers

Work title
Type: preprint

OpenAlex work type
Language: en

Primary language
Publication year: 2025

Year of publication
Publication date: 2025-09-26

Full publication date if available
Authors: Duc Kieu, Kien Do, Tuan Hoang, Thao Minh Le, Tung Kieu, Dang H. Nguyen, Thin Nguyen

List of authors in order
Landing page: https://arxiv.org/abs/2510.03252

Publisher landing page
PDF URL: https://arxiv.org/pdf/2510.03252

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/2510.03252

Direct OA link when available
Cited by: 0

Total citation count in OpenAlex

Full payload

id	https://openalex.org/W4414968558
doi	https://doi.org/10.48550/arxiv.2510.03252
ids.doi	https://doi.org/10.48550/arxiv.2510.03252
ids.openalex	https://openalex.org/W4414968558
fwci
type	preprint
title	Universal Multi-Domain Translation via Diffusion Routers
biblio.issue
biblio.volume
biblio.last_page
biblio.first_page
topics[0].id	https://openalex.org/T12326
topics[0].field.id	https://openalex.org/fields/17
topics[0].field.display_name	Computer Science
topics[0].score	0.9526000022888184
topics[0].domain.id	https://openalex.org/domains/3
topics[0].domain.display_name	Physical Sciences
topics[0].subfield.id	https://openalex.org/subfields/1708
topics[0].subfield.display_name	Hardware and Architecture
topics[0].display_name	Network Packet Processing and Optimization
is_xpac	False
apc_list
apc_paid
language	en
locations[0].id	pmh:oai:arXiv.org:2510.03252
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/2510.03252
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/2510.03252
locations[1].id	doi:10.48550/arxiv.2510.03252
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	cc-by
locations[1].pdf_url
locations[1].version
locations[1].raw_type	article
locations[1].license_id	https://openalex.org/licenses/cc-by
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.2510.03252
indexed_in	arxiv, datacite
authorships[0].author.id	https://openalex.org/A5093884421
authorships[0].author.orcid
authorships[0].author.display_name	Duc Kieu
authorships[0].author_position	first
authorships[0].raw_author_name	Kieu, Duc
authorships[0].is_corresponding	False
authorships[1].author.id	https://openalex.org/A5001806269
authorships[1].author.orcid	https://orcid.org/0000-0002-0119-122X
authorships[1].author.display_name	Kien Do
authorships[1].author_position	middle
authorships[1].raw_author_name	Do, Kien
authorships[1].is_corresponding	False
authorships[2].author.id	https://openalex.org/A5080253938
authorships[2].author.orcid	https://orcid.org/0000-0002-1076-8043
authorships[2].author.display_name	Tuan Hoang
authorships[2].author_position	middle
authorships[2].raw_author_name	Hoang, Tuan
authorships[2].is_corresponding	False
authorships[3].author.id	https://openalex.org/A5079045166
authorships[3].author.orcid	https://orcid.org/0000-0002-8089-9962
authorships[3].author.display_name	Thao Minh Le
authorships[3].author_position	middle
authorships[3].raw_author_name	Le, Thao Minh
authorships[3].is_corresponding	False
authorships[4].author.id	https://openalex.org/A5016799625
authorships[4].author.orcid	https://orcid.org/0000-0002-7696-1444
authorships[4].author.display_name	Tung Kieu
authorships[4].author_position	middle
authorships[4].raw_author_name	Kieu, Tung
authorships[4].is_corresponding	False
authorships[5].author.id	https://openalex.org/A5101752860
authorships[5].author.orcid	https://orcid.org/0000-0002-9007-8690
authorships[5].author.display_name	Dang H. Nguyen
authorships[5].author_position	middle
authorships[5].raw_author_name	Nguyen, Dang
authorships[5].is_corresponding	False
authorships[6].author.id	https://openalex.org/A5100705489
authorships[6].author.orcid	https://orcid.org/0000-0003-3467-8963
authorships[6].author.display_name	Thin Nguyen
authorships[6].author_position	last
authorships[6].raw_author_name	Nguyen, Thin
authorships[6].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/2510.03252
open_access.oa_status	green
open_access.any_repository_has_fulltext	False
created_date	2025-10-09T00:00:00
display_name	Universal Multi-Domain Translation via Diffusion Routers
has_fulltext	False
is_retracted	False
updated_date	2025-11-06T06:51:31.235846
primary_topic.id	https://openalex.org/T12326
primary_topic.field.id	https://openalex.org/fields/17
primary_topic.field.display_name	Computer Science
primary_topic.score	0.9526000022888184
primary_topic.domain.id	https://openalex.org/domains/3
primary_topic.domain.display_name	Physical Sciences
primary_topic.subfield.id	https://openalex.org/subfields/1708
primary_topic.subfield.display_name	Hardware and Architecture
primary_topic.display_name	Network Packet Processing and Optimization
cited_by_count	0
locations_count	2
best_oa_location.id	pmh:oai:arXiv.org:2510.03252
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/2510.03252
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/2510.03252
primary_location.id	pmh:oai:arXiv.org:2510.03252
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/2510.03252
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/2510.03252
publication_date	2025-09-26
publication_year	2025
referenced_works_count	0
abstract_inverted_index.a	40, 60, 72, 82, 108, 114, 161
abstract_inverted_index.DR	94, 135, 159
abstract_inverted_index.To	63
abstract_inverted_index.We	35, 105
abstract_inverted_index.an	118
abstract_inverted_index.as	154, 160
abstract_inverted_index.by	99
abstract_inverted_index.in	25
abstract_inverted_index.of	42, 51
abstract_inverted_index.on	87, 130
abstract_inverted_index.or	18
abstract_inverted_index.to	4, 46, 123
abstract_inverted_index.we	67
abstract_inverted_index.$K$	52
abstract_inverted_index.MDT	38, 43
abstract_inverted_index.all	78
abstract_inverted_index.and	30, 90, 117, 142, 149, 163
abstract_inverted_index.any	49
abstract_inverted_index.can	19
abstract_inverted_index.for	139, 166
abstract_inverted_index.the	88, 102
abstract_inverted_index.yet	10
abstract_inverted_index.UMDT	133
abstract_inverted_index.aims	3
abstract_inverted_index.both	140
abstract_inverted_index.cost	148
abstract_inverted_index.many	32
abstract_inverted_index.only	20, 55
abstract_inverted_index.pair	50
abstract_inverted_index.seen	24
abstract_inverted_index.such	153
abstract_inverted_index.that	44, 76
abstract_inverted_index.this	65
abstract_inverted_index.with	59, 81, 113
abstract_inverted_index.$K-1$	56
abstract_inverted_index.(DR),	71
abstract_inverted_index.(MDT)	2
abstract_inverted_index.These	156
abstract_inverted_index.fully	15
abstract_inverted_index.learn	5
abstract_inverted_index.noise	84
abstract_inverted_index.novel	109, 151
abstract_inverted_index.pairs	23
abstract_inverted_index.seeks	45
abstract_inverted_index.tasks	152
abstract_inverted_index.their	28
abstract_inverted_index.three	131
abstract_inverted_index.using	54
abstract_inverted_index.while	145
abstract_inverted_index.Router	70
abstract_inverted_index.across	169
abstract_inverted_index.direct	125, 143
abstract_inverted_index.domain	22, 92
abstract_inverted_index.either	13
abstract_inverted_index.handle	21
abstract_inverted_index.models	77
abstract_inverted_index.paired	57
abstract_inverted_index.single	83
abstract_inverted_index.source	89
abstract_inverted_index.tackle	64
abstract_inverted_index.target	91
abstract_inverted_index.tuples	17
abstract_inverted_index.(UMDT),	39
abstract_inverted_index.Through	128
abstract_inverted_index.Tweedie	120
abstract_inverted_index.aligned	16
abstract_inverted_index.between	7, 48
abstract_inverted_index.central	61, 103
abstract_inverted_index.domain.	62, 104
abstract_inverted_index.domains	53
abstract_inverted_index.enables	95
abstract_inverted_index.further	106
abstract_inverted_index.labels.	93
abstract_inverted_index.propose	68
abstract_inverted_index.require	14
abstract_inverted_index.results	138, 157
abstract_inverted_index.routing	100
abstract_inverted_index.support	124
abstract_inverted_index.through	101
abstract_inverted_index.unified	73
abstract_inverted_index.achieves	136
abstract_inverted_index.datasets	58
abstract_inverted_index.domains,	9
abstract_inverted_index.domains.	171
abstract_inverted_index.existing	11
abstract_inverted_index.indirect	96, 141
abstract_inverted_index.learning	111
abstract_inverted_index.limiting	27
abstract_inverted_index.lowering	146
abstract_inverted_index.multiple	8, 170
abstract_inverted_index.problem,	66
abstract_inverted_index.sampling	147
abstract_inverted_index.scalable	110, 162
abstract_inverted_index.strategy	112
abstract_inverted_index.Diffusion	69
abstract_inverted_index.efficient	119
abstract_inverted_index.establish	158
abstract_inverted_index.excluding	31
abstract_inverted_index.framework	75, 165
abstract_inverted_index.introduce	36, 107
abstract_inverted_index.mappings.	34, 127
abstract_inverted_index.objective	116
abstract_inverted_index.predictor	85
abstract_inverted_index.procedure	122
abstract_inverted_index.training,	26
abstract_inverted_index.translate	47
abstract_inverted_index.universal	37, 167
abstract_inverted_index.unlocking	150
abstract_inverted_index.versatile	164
abstract_inverted_index.approaches	12
abstract_inverted_index.evaluation	129
abstract_inverted_index.refinement	121
abstract_inverted_index.benchmarks,	134
abstract_inverted_index.conditioned	86
abstract_inverted_index.large-scale	132
abstract_inverted_index.non-central	97, 126
abstract_inverted_index.translation	1, 168
abstract_inverted_index.Multi-domain	0
abstract_inverted_index.cross-domain	33
abstract_inverted_index.practicality	29
abstract_inverted_index.translations	6, 80, 98
abstract_inverted_index.translations,	144
abstract_inverted_index.generalization	41
abstract_inverted_index.diffusion-based	74
abstract_inverted_index.state-of-the-art	137
abstract_inverted_index.variational-bound	115
abstract_inverted_index.central$\leftrightarrow$non-central	79
abstract_inverted_index.sketch$\leftrightarrow$segmentation.	155
cited_by_percentile_year
countries_distinct_count	0
institutions_distinct_count	7
citation_normalized_percentile