Learning When to Speak: Latency and Quality Trade-offs for Simultaneous Speech-to-Speech Translation with Offline Models Article Swipe
Liam Dugan
,
Anshul Wadhawan
,
Kyle Spence
,
Chris Callison-Burch
,
Morgan McGuire
,
Victor Zordan
·
YOU?
·
· 2023
· Open Access
·
· DOI: https://doi.org/10.48550/arxiv.2306.01201
YOU?
·
· 2023
· Open Access
·
· DOI: https://doi.org/10.48550/arxiv.2306.01201
Recent work in speech-to-speech translation (S2ST) has focused primarily on offline settings, where the full input utterance is available before any output is given. This, however, is not reasonable in many real-world scenarios. In latency-sensitive applications, rather than waiting for the full utterance, translations should be spoken as soon as the information in the input is present. In this work, we introduce a system for simultaneous S2ST targeting real-world use cases. Our system supports translation from 57 languages to English with tunable parameters for dynamically adjusting the latency of the output -- including four policies for determining when to speak an output sequence. We show that these policies achieve offline-level accuracy with minimal increases in latency over a Greedy (wait-$k$) baseline. We open-source our evaluation code and interactive test script to aid future SimulS2ST research and application development.
Related Topics
Concepts
Metadata
- Type
- preprint
- Language
- en
- Landing Page
- http://arxiv.org/abs/2306.01201
- https://arxiv.org/pdf/2306.01201
- OA Status
- green
- Related Works
- 10
- OpenAlex ID
- https://openalex.org/W4379471040
All OpenAlex metadata
Raw OpenAlex JSON
- OpenAlex ID
-
https://openalex.org/W4379471040Canonical identifier for this work in OpenAlex
- DOI
-
https://doi.org/10.48550/arxiv.2306.01201Digital Object Identifier
- Title
-
Learning When to Speak: Latency and Quality Trade-offs for Simultaneous Speech-to-Speech Translation with Offline ModelsWork title
- Type
-
preprintOpenAlex work type
- Language
-
enPrimary language
- Publication year
-
2023Year of publication
- Publication date
-
2023-06-01Full publication date if available
- Authors
-
Liam Dugan, Anshul Wadhawan, Kyle Spence, Chris Callison-Burch, Morgan McGuire, Victor ZordanList of authors in order
- Landing page
-
https://arxiv.org/abs/2306.01201Publisher landing page
- PDF URL
-
https://arxiv.org/pdf/2306.01201Direct link to full text PDF
- Open access
-
YesWhether a free full text is available
- OA status
-
greenOpen access status per OpenAlex
- OA URL
-
https://arxiv.org/pdf/2306.01201Direct OA link when available
- Concepts
-
Computer science, Speech translation, Utterance, Latency (audio), Speech recognition, Machine translation, Natural language processing, Artificial intelligence, TelecommunicationsTop concepts (fields/topics) attached by OpenAlex
- Cited by
-
0Total citation count in OpenAlex
- Related works (count)
-
10Other works algorithmically related by OpenAlex
Full payload
| id | https://openalex.org/W4379471040 |
|---|---|
| doi | https://doi.org/10.48550/arxiv.2306.01201 |
| ids.doi | https://doi.org/10.48550/arxiv.2306.01201 |
| ids.openalex | https://openalex.org/W4379471040 |
| fwci | |
| type | preprint |
| title | Learning When to Speak: Latency and Quality Trade-offs for Simultaneous Speech-to-Speech Translation with Offline Models |
| biblio.issue | |
| biblio.volume | |
| biblio.last_page | |
| biblio.first_page | |
| topics[0].id | https://openalex.org/T10181 |
| topics[0].field.id | https://openalex.org/fields/17 |
| topics[0].field.display_name | Computer Science |
| topics[0].score | 0.9987999796867371 |
| topics[0].domain.id | https://openalex.org/domains/3 |
| topics[0].domain.display_name | Physical Sciences |
| topics[0].subfield.id | https://openalex.org/subfields/1702 |
| topics[0].subfield.display_name | Artificial Intelligence |
| topics[0].display_name | Natural Language Processing Techniques |
| topics[1].id | https://openalex.org/T12031 |
| topics[1].field.id | https://openalex.org/fields/17 |
| topics[1].field.display_name | Computer Science |
| topics[1].score | 0.9987000226974487 |
| topics[1].domain.id | https://openalex.org/domains/3 |
| topics[1].domain.display_name | Physical Sciences |
| topics[1].subfield.id | https://openalex.org/subfields/1702 |
| topics[1].subfield.display_name | Artificial Intelligence |
| topics[1].display_name | Speech and dialogue systems |
| topics[2].id | https://openalex.org/T10201 |
| topics[2].field.id | https://openalex.org/fields/17 |
| topics[2].field.display_name | Computer Science |
| topics[2].score | 0.9936000108718872 |
| topics[2].domain.id | https://openalex.org/domains/3 |
| topics[2].domain.display_name | Physical Sciences |
| topics[2].subfield.id | https://openalex.org/subfields/1702 |
| topics[2].subfield.display_name | Artificial Intelligence |
| topics[2].display_name | Speech Recognition and Synthesis |
| is_xpac | False |
| apc_list | |
| apc_paid | |
| concepts[0].id | https://openalex.org/C41008148 |
| concepts[0].level | 0 |
| concepts[0].score | 0.822070300579071 |
| concepts[0].wikidata | https://www.wikidata.org/wiki/Q21198 |
| concepts[0].display_name | Computer science |
| concepts[1].id | https://openalex.org/C2780366754 |
| concepts[1].level | 3 |
| concepts[1].score | 0.8023918271064758 |
| concepts[1].wikidata | https://www.wikidata.org/wiki/Q7494857 |
| concepts[1].display_name | Speech translation |
| concepts[2].id | https://openalex.org/C2775852435 |
| concepts[2].level | 2 |
| concepts[2].score | 0.7788740396499634 |
| concepts[2].wikidata | https://www.wikidata.org/wiki/Q258403 |
| concepts[2].display_name | Utterance |
| concepts[3].id | https://openalex.org/C82876162 |
| concepts[3].level | 2 |
| concepts[3].score | 0.7761354446411133 |
| concepts[3].wikidata | https://www.wikidata.org/wiki/Q17096504 |
| concepts[3].display_name | Latency (audio) |
| concepts[4].id | https://openalex.org/C28490314 |
| concepts[4].level | 1 |
| concepts[4].score | 0.6473566293716431 |
| concepts[4].wikidata | https://www.wikidata.org/wiki/Q189436 |
| concepts[4].display_name | Speech recognition |
| concepts[5].id | https://openalex.org/C203005215 |
| concepts[5].level | 2 |
| concepts[5].score | 0.38813143968582153 |
| concepts[5].wikidata | https://www.wikidata.org/wiki/Q79798 |
| concepts[5].display_name | Machine translation |
| concepts[6].id | https://openalex.org/C204321447 |
| concepts[6].level | 1 |
| concepts[6].score | 0.3728829026222229 |
| concepts[6].wikidata | https://www.wikidata.org/wiki/Q30642 |
| concepts[6].display_name | Natural language processing |
| concepts[7].id | https://openalex.org/C154945302 |
| concepts[7].level | 1 |
| concepts[7].score | 0.36504876613616943 |
| concepts[7].wikidata | https://www.wikidata.org/wiki/Q11660 |
| concepts[7].display_name | Artificial intelligence |
| concepts[8].id | https://openalex.org/C76155785 |
| concepts[8].level | 1 |
| concepts[8].score | 0.08432206511497498 |
| concepts[8].wikidata | https://www.wikidata.org/wiki/Q418 |
| concepts[8].display_name | Telecommunications |
| keywords[0].id | https://openalex.org/keywords/computer-science |
| keywords[0].score | 0.822070300579071 |
| keywords[0].display_name | Computer science |
| keywords[1].id | https://openalex.org/keywords/speech-translation |
| keywords[1].score | 0.8023918271064758 |
| keywords[1].display_name | Speech translation |
| keywords[2].id | https://openalex.org/keywords/utterance |
| keywords[2].score | 0.7788740396499634 |
| keywords[2].display_name | Utterance |
| keywords[3].id | https://openalex.org/keywords/latency |
| keywords[3].score | 0.7761354446411133 |
| keywords[3].display_name | Latency (audio) |
| keywords[4].id | https://openalex.org/keywords/speech-recognition |
| keywords[4].score | 0.6473566293716431 |
| keywords[4].display_name | Speech recognition |
| keywords[5].id | https://openalex.org/keywords/machine-translation |
| keywords[5].score | 0.38813143968582153 |
| keywords[5].display_name | Machine translation |
| keywords[6].id | https://openalex.org/keywords/natural-language-processing |
| keywords[6].score | 0.3728829026222229 |
| keywords[6].display_name | Natural language processing |
| keywords[7].id | https://openalex.org/keywords/artificial-intelligence |
| keywords[7].score | 0.36504876613616943 |
| keywords[7].display_name | Artificial intelligence |
| keywords[8].id | https://openalex.org/keywords/telecommunications |
| keywords[8].score | 0.08432206511497498 |
| keywords[8].display_name | Telecommunications |
| language | en |
| locations[0].id | pmh:oai:arXiv.org:2306.01201 |
| 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/2306.01201 |
| locations[0].version | submittedVersion |
| locations[0].raw_type | |
| 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/2306.01201 |
| locations[1].id | doi:10.48550/arxiv.2306.01201 |
| 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.2306.01201 |
| indexed_in | arxiv, datacite |
| authorships[0].author.id | https://openalex.org/A5059969678 |
| authorships[0].author.orcid | https://orcid.org/0000-0001-6350-5944 |
| authorships[0].author.display_name | Liam Dugan |
| authorships[0].author_position | first |
| authorships[0].raw_author_name | Dugan, Liam |
| authorships[0].is_corresponding | False |
| authorships[1].author.id | https://openalex.org/A5069978153 |
| authorships[1].author.orcid | |
| authorships[1].author.display_name | Anshul Wadhawan |
| authorships[1].author_position | middle |
| authorships[1].raw_author_name | Wadhawan, Anshul |
| authorships[1].is_corresponding | False |
| authorships[2].author.id | https://openalex.org/A5104330009 |
| authorships[2].author.orcid | |
| authorships[2].author.display_name | Kyle Spence |
| authorships[2].author_position | middle |
| authorships[2].raw_author_name | Spence, Kyle |
| authorships[2].is_corresponding | False |
| authorships[3].author.id | https://openalex.org/A5068508539 |
| authorships[3].author.orcid | https://orcid.org/0000-0001-8196-1943 |
| authorships[3].author.display_name | Chris Callison-Burch |
| authorships[3].author_position | middle |
| authorships[3].raw_author_name | Callison-Burch, Chris |
| authorships[3].is_corresponding | False |
| authorships[4].author.id | https://openalex.org/A5101915674 |
| authorships[4].author.orcid | https://orcid.org/0000-0003-1074-0953 |
| authorships[4].author.display_name | Morgan McGuire |
| authorships[4].author_position | middle |
| authorships[4].raw_author_name | McGuire, Morgan |
| authorships[4].is_corresponding | False |
| authorships[5].author.id | https://openalex.org/A5000237739 |
| authorships[5].author.orcid | https://orcid.org/0000-0002-7309-7013 |
| authorships[5].author.display_name | Victor Zordan |
| authorships[5].author_position | last |
| authorships[5].raw_author_name | Zordan, Victor |
| authorships[5].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/2306.01201 |
| open_access.oa_status | green |
| open_access.any_repository_has_fulltext | False |
| created_date | 2023-06-07T00:00:00 |
| display_name | Learning When to Speak: Latency and Quality Trade-offs for Simultaneous Speech-to-Speech Translation with Offline Models |
| has_fulltext | False |
| is_retracted | False |
| updated_date | 2025-11-06T06:51:31.235846 |
| primary_topic.id | https://openalex.org/T10181 |
| primary_topic.field.id | https://openalex.org/fields/17 |
| primary_topic.field.display_name | Computer Science |
| primary_topic.score | 0.9987999796867371 |
| primary_topic.domain.id | https://openalex.org/domains/3 |
| primary_topic.domain.display_name | Physical Sciences |
| primary_topic.subfield.id | https://openalex.org/subfields/1702 |
| primary_topic.subfield.display_name | Artificial Intelligence |
| primary_topic.display_name | Natural Language Processing Techniques |
| related_works | https://openalex.org/W2529301793, https://openalex.org/W2384121599, https://openalex.org/W2038083449, https://openalex.org/W3177678247, https://openalex.org/W2333799855, https://openalex.org/W1999617572, https://openalex.org/W2944572343, https://openalex.org/W2351687372, https://openalex.org/W2004087835, https://openalex.org/W4223977554 |
| cited_by_count | 0 |
| locations_count | 2 |
| best_oa_location.id | pmh:oai:arXiv.org:2306.01201 |
| 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/2306.01201 |
| best_oa_location.version | submittedVersion |
| best_oa_location.raw_type | |
| 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/2306.01201 |
| primary_location.id | pmh:oai:arXiv.org:2306.01201 |
| 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/2306.01201 |
| primary_location.version | submittedVersion |
| primary_location.raw_type | |
| 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/2306.01201 |
| publication_date | 2023-06-01 |
| publication_year | 2023 |
| referenced_works_count | 0 |
| abstract_inverted_index.a | 62, 117 |
| abstract_inverted_index.-- | 91 |
| abstract_inverted_index.57 | 76 |
| abstract_inverted_index.In | 33, 57 |
| abstract_inverted_index.We | 103, 121 |
| abstract_inverted_index.an | 100 |
| abstract_inverted_index.as | 47, 49 |
| abstract_inverted_index.be | 45 |
| abstract_inverted_index.in | 2, 29, 52, 114 |
| abstract_inverted_index.is | 17, 22, 26, 55 |
| abstract_inverted_index.of | 88 |
| abstract_inverted_index.on | 9 |
| abstract_inverted_index.to | 78, 98, 130 |
| abstract_inverted_index.we | 60 |
| abstract_inverted_index.Our | 71 |
| abstract_inverted_index.aid | 131 |
| abstract_inverted_index.and | 126, 135 |
| abstract_inverted_index.any | 20 |
| abstract_inverted_index.for | 39, 64, 83, 95 |
| abstract_inverted_index.has | 6 |
| abstract_inverted_index.not | 27 |
| abstract_inverted_index.our | 123 |
| abstract_inverted_index.the | 13, 40, 50, 53, 86, 89 |
| abstract_inverted_index.use | 69 |
| abstract_inverted_index.S2ST | 66 |
| abstract_inverted_index.code | 125 |
| abstract_inverted_index.four | 93 |
| abstract_inverted_index.from | 75 |
| abstract_inverted_index.full | 14, 41 |
| abstract_inverted_index.many | 30 |
| abstract_inverted_index.over | 116 |
| abstract_inverted_index.show | 104 |
| abstract_inverted_index.soon | 48 |
| abstract_inverted_index.test | 128 |
| abstract_inverted_index.than | 37 |
| abstract_inverted_index.that | 105 |
| abstract_inverted_index.this | 58 |
| abstract_inverted_index.when | 97 |
| abstract_inverted_index.with | 80, 111 |
| abstract_inverted_index.work | 1 |
| abstract_inverted_index.This, | 24 |
| abstract_inverted_index.input | 15, 54 |
| abstract_inverted_index.speak | 99 |
| abstract_inverted_index.these | 106 |
| abstract_inverted_index.where | 12 |
| abstract_inverted_index.work, | 59 |
| abstract_inverted_index.(S2ST) | 5 |
| abstract_inverted_index.Greedy | 118 |
| abstract_inverted_index.Recent | 0 |
| abstract_inverted_index.before | 19 |
| abstract_inverted_index.cases. | 70 |
| abstract_inverted_index.future | 132 |
| abstract_inverted_index.given. | 23 |
| abstract_inverted_index.output | 21, 90, 101 |
| abstract_inverted_index.rather | 36 |
| abstract_inverted_index.script | 129 |
| abstract_inverted_index.should | 44 |
| abstract_inverted_index.spoken | 46 |
| abstract_inverted_index.system | 63, 72 |
| abstract_inverted_index.English | 79 |
| abstract_inverted_index.achieve | 108 |
| abstract_inverted_index.focused | 7 |
| abstract_inverted_index.latency | 87, 115 |
| abstract_inverted_index.minimal | 112 |
| abstract_inverted_index.offline | 10 |
| abstract_inverted_index.tunable | 81 |
| abstract_inverted_index.waiting | 38 |
| abstract_inverted_index.accuracy | 110 |
| abstract_inverted_index.however, | 25 |
| abstract_inverted_index.policies | 94, 107 |
| abstract_inverted_index.present. | 56 |
| abstract_inverted_index.research | 134 |
| abstract_inverted_index.supports | 73 |
| abstract_inverted_index.SimulS2ST | 133 |
| abstract_inverted_index.adjusting | 85 |
| abstract_inverted_index.available | 18 |
| abstract_inverted_index.baseline. | 120 |
| abstract_inverted_index.including | 92 |
| abstract_inverted_index.increases | 113 |
| abstract_inverted_index.introduce | 61 |
| abstract_inverted_index.languages | 77 |
| abstract_inverted_index.primarily | 8 |
| abstract_inverted_index.sequence. | 102 |
| abstract_inverted_index.settings, | 11 |
| abstract_inverted_index.targeting | 67 |
| abstract_inverted_index.utterance | 16 |
| abstract_inverted_index.(wait-$k$) | 119 |
| abstract_inverted_index.evaluation | 124 |
| abstract_inverted_index.parameters | 82 |
| abstract_inverted_index.real-world | 31, 68 |
| abstract_inverted_index.reasonable | 28 |
| abstract_inverted_index.scenarios. | 32 |
| abstract_inverted_index.utterance, | 42 |
| abstract_inverted_index.application | 136 |
| abstract_inverted_index.determining | 96 |
| abstract_inverted_index.dynamically | 84 |
| abstract_inverted_index.information | 51 |
| abstract_inverted_index.interactive | 127 |
| abstract_inverted_index.open-source | 122 |
| abstract_inverted_index.translation | 4, 74 |
| abstract_inverted_index.development. | 137 |
| abstract_inverted_index.simultaneous | 65 |
| abstract_inverted_index.translations | 43 |
| abstract_inverted_index.applications, | 35 |
| abstract_inverted_index.offline-level | 109 |
| abstract_inverted_index.speech-to-speech | 3 |
| abstract_inverted_index.latency-sensitive | 34 |
| cited_by_percentile_year | |
| countries_distinct_count | 0 |
| institutions_distinct_count | 6 |
| sustainable_development_goals[0].id | https://metadata.un.org/sdg/17 |
| sustainable_development_goals[0].score | 0.4300000071525574 |
| sustainable_development_goals[0].display_name | Partnerships for the goals |
| citation_normalized_percentile |