Long-Tailed Object Detection Pre-training: Dynamic Rebalancing Contrastive Learning with Dual Reconstruction Article Swipe
Chenlong Duan
,
Yongping Li
,
Xiu-Shen Wei
,
Lin Zhao
·
YOU?
·
· 2024
· Open Access
·
· DOI: https://doi.org/10.48550/arxiv.2411.09453
YOU?
·
· 2024
· Open Access
·
· DOI: https://doi.org/10.48550/arxiv.2411.09453
Pre-training plays a vital role in various vision tasks, such as object recognition and detection. Commonly used pre-training methods, which typically rely on randomized approaches like uniform or Gaussian distributions to initialize model parameters, often fall short when confronted with long-tailed distributions, especially in detection tasks. This is largely due to extreme data imbalance and the issue of simplicity bias. In this paper, we introduce a novel pre-training framework for object detection, called Dynamic Rebalancing Contrastive Learning with Dual Reconstruction (2DRCL). Our method builds on a Holistic-Local Contrastive Learning mechanism, which aligns pre-training with object detection by capturing both global contextual semantics and detailed local patterns. To tackle the imbalance inherent in long-tailed data, we design a dynamic rebalancing strategy that adjusts the sampling of underrepresented instances throughout the pre-training process, ensuring better representation of tail classes. Moreover, Dual Reconstruction addresses simplicity bias by enforcing a reconstruction task aligned with the self-consistency principle, specifically benefiting underrepresented tail classes. Experiments on COCO and LVIS v1.0 datasets demonstrate the effectiveness of our method, particularly in improving the mAP/AP scores for tail classes.
Related Topics
Concepts
Metadata
- Type
- preprint
- Language
- en
- Landing Page
- http://arxiv.org/abs/2411.09453
- https://arxiv.org/pdf/2411.09453
- OA Status
- green
- Related Works
- 10
- OpenAlex ID
- https://openalex.org/W4404450909
All OpenAlex metadata
Raw OpenAlex JSON
- OpenAlex ID
-
https://openalex.org/W4404450909Canonical identifier for this work in OpenAlex
- DOI
-
https://doi.org/10.48550/arxiv.2411.09453Digital Object Identifier
- Title
-
Long-Tailed Object Detection Pre-training: Dynamic Rebalancing Contrastive Learning with Dual ReconstructionWork title
- Type
-
preprintOpenAlex work type
- Language
-
enPrimary language
- Publication year
-
2024Year of publication
- Publication date
-
2024-11-14Full publication date if available
- Authors
-
Chenlong Duan, Yongping Li, Xiu-Shen Wei, Lin ZhaoList of authors in order
- Landing page
-
https://arxiv.org/abs/2411.09453Publisher landing page
- PDF URL
-
https://arxiv.org/pdf/2411.09453Direct 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/2411.09453Direct OA link when available
- Concepts
-
Dual (grammatical number), Training (meteorology), Computer science, Object (grammar), Artificial intelligence, Computer vision, Geography, Linguistics, Meteorology, PhilosophyTop 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/W4404450909 |
|---|---|
| doi | https://doi.org/10.48550/arxiv.2411.09453 |
| ids.doi | https://doi.org/10.48550/arxiv.2411.09453 |
| ids.openalex | https://openalex.org/W4404450909 |
| fwci | |
| type | preprint |
| title | Long-Tailed Object Detection Pre-training: Dynamic Rebalancing Contrastive Learning with Dual Reconstruction |
| biblio.issue | |
| biblio.volume | |
| biblio.last_page | |
| biblio.first_page | |
| topics[0].id | https://openalex.org/T11307 |
| topics[0].field.id | https://openalex.org/fields/17 |
| topics[0].field.display_name | Computer Science |
| topics[0].score | 0.8607000112533569 |
| 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 | Domain Adaptation and Few-Shot Learning |
| topics[1].id | https://openalex.org/T10036 |
| topics[1].field.id | https://openalex.org/fields/17 |
| topics[1].field.display_name | Computer Science |
| topics[1].score | 0.7322999835014343 |
| topics[1].domain.id | https://openalex.org/domains/3 |
| topics[1].domain.display_name | Physical Sciences |
| topics[1].subfield.id | https://openalex.org/subfields/1707 |
| topics[1].subfield.display_name | Computer Vision and Pattern Recognition |
| topics[1].display_name | Advanced Neural Network Applications |
| topics[2].id | https://openalex.org/T12676 |
| topics[2].field.id | https://openalex.org/fields/17 |
| topics[2].field.display_name | Computer Science |
| topics[2].score | 0.7059999704360962 |
| 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 | Machine Learning and ELM |
| is_xpac | False |
| apc_list | |
| apc_paid | |
| concepts[0].id | https://openalex.org/C2780980858 |
| concepts[0].level | 2 |
| concepts[0].score | 0.7679574489593506 |
| concepts[0].wikidata | https://www.wikidata.org/wiki/Q110022 |
| concepts[0].display_name | Dual (grammatical number) |
| concepts[1].id | https://openalex.org/C2777211547 |
| concepts[1].level | 2 |
| concepts[1].score | 0.6806340217590332 |
| concepts[1].wikidata | https://www.wikidata.org/wiki/Q17141490 |
| concepts[1].display_name | Training (meteorology) |
| concepts[2].id | https://openalex.org/C41008148 |
| concepts[2].level | 0 |
| concepts[2].score | 0.6076586246490479 |
| concepts[2].wikidata | https://www.wikidata.org/wiki/Q21198 |
| concepts[2].display_name | Computer science |
| concepts[3].id | https://openalex.org/C2781238097 |
| concepts[3].level | 2 |
| concepts[3].score | 0.5586576461791992 |
| concepts[3].wikidata | https://www.wikidata.org/wiki/Q175026 |
| concepts[3].display_name | Object (grammar) |
| concepts[4].id | https://openalex.org/C154945302 |
| concepts[4].level | 1 |
| concepts[4].score | 0.5411760807037354 |
| concepts[4].wikidata | https://www.wikidata.org/wiki/Q11660 |
| concepts[4].display_name | Artificial intelligence |
| concepts[5].id | https://openalex.org/C31972630 |
| concepts[5].level | 1 |
| concepts[5].score | 0.3835001587867737 |
| concepts[5].wikidata | https://www.wikidata.org/wiki/Q844240 |
| concepts[5].display_name | Computer vision |
| concepts[6].id | https://openalex.org/C205649164 |
| concepts[6].level | 0 |
| concepts[6].score | 0.10482704639434814 |
| concepts[6].wikidata | https://www.wikidata.org/wiki/Q1071 |
| concepts[6].display_name | Geography |
| concepts[7].id | https://openalex.org/C41895202 |
| concepts[7].level | 1 |
| concepts[7].score | 0.09066051244735718 |
| concepts[7].wikidata | https://www.wikidata.org/wiki/Q8162 |
| concepts[7].display_name | Linguistics |
| concepts[8].id | https://openalex.org/C153294291 |
| concepts[8].level | 1 |
| concepts[8].score | 0.0 |
| concepts[8].wikidata | https://www.wikidata.org/wiki/Q25261 |
| concepts[8].display_name | Meteorology |
| concepts[9].id | https://openalex.org/C138885662 |
| concepts[9].level | 0 |
| concepts[9].score | 0.0 |
| concepts[9].wikidata | https://www.wikidata.org/wiki/Q5891 |
| concepts[9].display_name | Philosophy |
| keywords[0].id | https://openalex.org/keywords/dual |
| keywords[0].score | 0.7679574489593506 |
| keywords[0].display_name | Dual (grammatical number) |
| keywords[1].id | https://openalex.org/keywords/training |
| keywords[1].score | 0.6806340217590332 |
| keywords[1].display_name | Training (meteorology) |
| keywords[2].id | https://openalex.org/keywords/computer-science |
| keywords[2].score | 0.6076586246490479 |
| keywords[2].display_name | Computer science |
| keywords[3].id | https://openalex.org/keywords/object |
| keywords[3].score | 0.5586576461791992 |
| keywords[3].display_name | Object (grammar) |
| keywords[4].id | https://openalex.org/keywords/artificial-intelligence |
| keywords[4].score | 0.5411760807037354 |
| keywords[4].display_name | Artificial intelligence |
| keywords[5].id | https://openalex.org/keywords/computer-vision |
| keywords[5].score | 0.3835001587867737 |
| keywords[5].display_name | Computer vision |
| keywords[6].id | https://openalex.org/keywords/geography |
| keywords[6].score | 0.10482704639434814 |
| keywords[6].display_name | Geography |
| keywords[7].id | https://openalex.org/keywords/linguistics |
| keywords[7].score | 0.09066051244735718 |
| keywords[7].display_name | Linguistics |
| language | en |
| locations[0].id | pmh:oai:arXiv.org:2411.09453 |
| 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/2411.09453 |
| 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/2411.09453 |
| locations[1].id | doi:10.48550/arxiv.2411.09453 |
| 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.2411.09453 |
| indexed_in | arxiv, datacite |
| authorships[0].author.id | https://openalex.org/A5100754598 |
| authorships[0].author.orcid | https://orcid.org/0000-0002-8093-6719 |
| authorships[0].author.display_name | Chenlong Duan |
| authorships[0].author_position | first |
| authorships[0].raw_author_name | Duan, Chen-Long |
| authorships[0].is_corresponding | False |
| authorships[1].author.id | https://openalex.org/A5100720497 |
| authorships[1].author.orcid | https://orcid.org/0000-0002-3253-4088 |
| authorships[1].author.display_name | Yongping Li |
| authorships[1].author_position | middle |
| authorships[1].raw_author_name | Li, Yong |
| authorships[1].is_corresponding | False |
| authorships[2].author.id | https://openalex.org/A5066964304 |
| authorships[2].author.orcid | https://orcid.org/0000-0002-8200-1845 |
| authorships[2].author.display_name | Xiu-Shen Wei |
| authorships[2].author_position | middle |
| authorships[2].raw_author_name | Wei, Xiu-Shen |
| authorships[2].is_corresponding | False |
| authorships[3].author.id | https://openalex.org/A5100351935 |
| authorships[3].author.orcid | https://orcid.org/0000-0003-2567-627X |
| authorships[3].author.display_name | Lin Zhao |
| authorships[3].author_position | last |
| authorships[3].raw_author_name | Zhao, Lin |
| authorships[3].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/2411.09453 |
| open_access.oa_status | green |
| open_access.any_repository_has_fulltext | False |
| created_date | 2024-11-17T00:00:00 |
| display_name | Long-Tailed Object Detection Pre-training: Dynamic Rebalancing Contrastive Learning with Dual Reconstruction |
| has_fulltext | False |
| is_retracted | False |
| updated_date | 2025-11-06T06:51:31.235846 |
| primary_topic.id | https://openalex.org/T11307 |
| primary_topic.field.id | https://openalex.org/fields/17 |
| primary_topic.field.display_name | Computer Science |
| primary_topic.score | 0.8607000112533569 |
| 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 | Domain Adaptation and Few-Shot Learning |
| related_works | https://openalex.org/W2772917594, https://openalex.org/W2036807459, https://openalex.org/W2058170566, https://openalex.org/W2755342338, https://openalex.org/W2166024367, https://openalex.org/W3116076068, https://openalex.org/W2229312674, https://openalex.org/W2951359407, https://openalex.org/W2079911747, https://openalex.org/W1969923398 |
| cited_by_count | 0 |
| locations_count | 2 |
| best_oa_location.id | pmh:oai:arXiv.org:2411.09453 |
| 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/2411.09453 |
| 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/2411.09453 |
| primary_location.id | pmh:oai:arXiv.org:2411.09453 |
| 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/2411.09453 |
| 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/2411.09453 |
| publication_date | 2024-11-14 |
| publication_year | 2024 |
| referenced_works_count | 0 |
| abstract_inverted_index.a | 2, 65, 85, 116, 145 |
| abstract_inverted_index.In | 60 |
| abstract_inverted_index.To | 106 |
| abstract_inverted_index.as | 10 |
| abstract_inverted_index.by | 96, 143 |
| abstract_inverted_index.in | 5, 43, 111, 172 |
| abstract_inverted_index.is | 47 |
| abstract_inverted_index.of | 57, 124, 134, 168 |
| abstract_inverted_index.on | 22, 84, 159 |
| abstract_inverted_index.or | 27 |
| abstract_inverted_index.to | 30, 50 |
| abstract_inverted_index.we | 63, 114 |
| abstract_inverted_index.Our | 81 |
| abstract_inverted_index.and | 13, 54, 102, 161 |
| abstract_inverted_index.due | 49 |
| abstract_inverted_index.for | 69, 177 |
| abstract_inverted_index.our | 169 |
| abstract_inverted_index.the | 55, 108, 122, 128, 150, 166, 174 |
| abstract_inverted_index.COCO | 160 |
| abstract_inverted_index.Dual | 78, 138 |
| abstract_inverted_index.LVIS | 162 |
| abstract_inverted_index.This | 46 |
| abstract_inverted_index.bias | 142 |
| abstract_inverted_index.both | 98 |
| abstract_inverted_index.data | 52 |
| abstract_inverted_index.fall | 35 |
| abstract_inverted_index.like | 25 |
| abstract_inverted_index.rely | 21 |
| abstract_inverted_index.role | 4 |
| abstract_inverted_index.such | 9 |
| abstract_inverted_index.tail | 135, 156, 178 |
| abstract_inverted_index.task | 147 |
| abstract_inverted_index.that | 120 |
| abstract_inverted_index.this | 61 |
| abstract_inverted_index.used | 16 |
| abstract_inverted_index.v1.0 | 163 |
| abstract_inverted_index.when | 37 |
| abstract_inverted_index.with | 39, 77, 93, 149 |
| abstract_inverted_index.bias. | 59 |
| abstract_inverted_index.data, | 113 |
| abstract_inverted_index.issue | 56 |
| abstract_inverted_index.local | 104 |
| abstract_inverted_index.model | 32 |
| abstract_inverted_index.novel | 66 |
| abstract_inverted_index.often | 34 |
| abstract_inverted_index.plays | 1 |
| abstract_inverted_index.short | 36 |
| abstract_inverted_index.vital | 3 |
| abstract_inverted_index.which | 19, 90 |
| abstract_inverted_index.aligns | 91 |
| abstract_inverted_index.better | 132 |
| abstract_inverted_index.builds | 83 |
| abstract_inverted_index.called | 72 |
| abstract_inverted_index.design | 115 |
| abstract_inverted_index.global | 99 |
| abstract_inverted_index.mAP/AP | 175 |
| abstract_inverted_index.method | 82 |
| abstract_inverted_index.object | 11, 70, 94 |
| abstract_inverted_index.paper, | 62 |
| abstract_inverted_index.scores | 176 |
| abstract_inverted_index.tackle | 107 |
| abstract_inverted_index.tasks, | 8 |
| abstract_inverted_index.tasks. | 45 |
| abstract_inverted_index.vision | 7 |
| abstract_inverted_index.Dynamic | 73 |
| abstract_inverted_index.adjusts | 121 |
| abstract_inverted_index.aligned | 148 |
| abstract_inverted_index.dynamic | 117 |
| abstract_inverted_index.extreme | 51 |
| abstract_inverted_index.largely | 48 |
| abstract_inverted_index.method, | 170 |
| abstract_inverted_index.uniform | 26 |
| abstract_inverted_index.various | 6 |
| abstract_inverted_index.(2DRCL). | 80 |
| abstract_inverted_index.Commonly | 15 |
| abstract_inverted_index.Gaussian | 28 |
| abstract_inverted_index.Learning | 76, 88 |
| abstract_inverted_index.classes. | 136, 157, 179 |
| abstract_inverted_index.datasets | 164 |
| abstract_inverted_index.detailed | 103 |
| abstract_inverted_index.ensuring | 131 |
| abstract_inverted_index.inherent | 110 |
| abstract_inverted_index.methods, | 18 |
| abstract_inverted_index.process, | 130 |
| abstract_inverted_index.sampling | 123 |
| abstract_inverted_index.strategy | 119 |
| abstract_inverted_index.Moreover, | 137 |
| abstract_inverted_index.addresses | 140 |
| abstract_inverted_index.capturing | 97 |
| abstract_inverted_index.detection | 44, 95 |
| abstract_inverted_index.enforcing | 144 |
| abstract_inverted_index.framework | 68 |
| abstract_inverted_index.imbalance | 53, 109 |
| abstract_inverted_index.improving | 173 |
| abstract_inverted_index.instances | 126 |
| abstract_inverted_index.introduce | 64 |
| abstract_inverted_index.patterns. | 105 |
| abstract_inverted_index.semantics | 101 |
| abstract_inverted_index.typically | 20 |
| abstract_inverted_index.approaches | 24 |
| abstract_inverted_index.benefiting | 154 |
| abstract_inverted_index.confronted | 38 |
| abstract_inverted_index.contextual | 100 |
| abstract_inverted_index.detection, | 71 |
| abstract_inverted_index.detection. | 14 |
| abstract_inverted_index.especially | 42 |
| abstract_inverted_index.initialize | 31 |
| abstract_inverted_index.mechanism, | 89 |
| abstract_inverted_index.principle, | 152 |
| abstract_inverted_index.randomized | 23 |
| abstract_inverted_index.simplicity | 58, 141 |
| abstract_inverted_index.throughout | 127 |
| abstract_inverted_index.Contrastive | 75, 87 |
| abstract_inverted_index.Experiments | 158 |
| abstract_inverted_index.Rebalancing | 74 |
| abstract_inverted_index.demonstrate | 165 |
| abstract_inverted_index.long-tailed | 40, 112 |
| abstract_inverted_index.parameters, | 33 |
| abstract_inverted_index.rebalancing | 118 |
| abstract_inverted_index.recognition | 12 |
| abstract_inverted_index.Pre-training | 0 |
| abstract_inverted_index.particularly | 171 |
| abstract_inverted_index.pre-training | 17, 67, 92, 129 |
| abstract_inverted_index.specifically | 153 |
| abstract_inverted_index.distributions | 29 |
| abstract_inverted_index.effectiveness | 167 |
| abstract_inverted_index.Holistic-Local | 86 |
| abstract_inverted_index.Reconstruction | 79, 139 |
| abstract_inverted_index.distributions, | 41 |
| abstract_inverted_index.reconstruction | 146 |
| abstract_inverted_index.representation | 133 |
| abstract_inverted_index.self-consistency | 151 |
| abstract_inverted_index.underrepresented | 125, 155 |
| cited_by_percentile_year | |
| countries_distinct_count | 0 |
| institutions_distinct_count | 4 |
| citation_normalized_percentile |