Smaller But Better: Unifying Layout Generation with Smaller Large Language Models.
Saved in:
| Title: | Smaller But Better: Unifying Layout Generation with Smaller Large Language Models. |
|---|---|
| Authors: | Zhang, Peirong1 (AUTHOR) eeprzhang@mail.scut.edu.cn, Zhang, Jiaxin1 (AUTHOR) msjxzhang@mail.scut.edu.cn, Cao, Jiahuan1 (AUTHOR) eejiahuancao@mail.scut.edu.cn, Li, Hongliang1 (AUTHOR) eehongliangli@mail.scut.edu.cn, Jin, Lianwen1 (AUTHOR) eelwjin@scut.edu.cn |
| Source: | International Journal of Computer Vision. Jul2025, Vol. 133 Issue 7, p3891-3917. 27p. |
| Subjects: | Language models, Complex variables, Encoding |
| Abstract: | We propose LGGPT, an LLM-based model tailored for unified layout generation. First, we propose Arbitrary Layout Instruction (ALI) and Universal Layout Response (ULR) as the uniform I/O template. ALI accommodates arbitrary layout generation task inputs across multiple layout domains, enabling LGGPT to unify both task-generic and domain-generic layout generation hitherto unexplored. Collectively, ALI and ULR boast a succinct structure that forgoes superfluous tokens typically found in existing HTML-based formats, facilitating efficient instruction tuning and boosting unified generation performance. In addition, we propose an Interval Quantization Encoding (IQE) strategy that compresses ALI into a more condensed structure. IQE precisely preserves valid layout clues while eliminating the less informative placeholders, facilitating LGGPT to capture complex and variable layout generation conditions during the unified training process. Experimental results demonstrate that LGGPT achieves superior or on par performance compared to existing methods. Notably, LGGPT strikes a prominent balance between proficiency and efficiency with a compact 1.5B parameter LLM, which beats prior 7B or 175B models even in the most extensive and challenging unified scenario. Furthermore, we underscore the necessity of employing LLMs for unified layout generation and suggest that 1.5B could be an optimal parameter size by comparing LLMs of varying scales. Code is available at https://github.com/NiceRingNode/LGGPT. [ABSTRACT FROM AUTHOR] |
| Copyright of International Journal of Computer Vision is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Engineering Source |
|
Full text is not displayed to guests.
Login for full access.
|
|
| FullText | Links: – Type: pdflink Text: Availability: 1 |
|---|---|
| Header | DbId: egs DbLabel: Engineering Source An: 185781607 AccessLevel: 6 PubType: Academic Journal PubTypeId: academicJournal PreciseRelevancyScore: 0 |
| IllustrationInfo | |
| Items | – Name: Title Label: Title Group: Ti Data: Smaller But Better: Unifying Layout Generation with Smaller Large Language Models. – Name: Author Label: Authors Group: Au Data: <searchLink fieldCode="AR" term="%22Zhang%2C+Peirong%22">Zhang, Peirong</searchLink><relatesTo>1</relatesTo> (AUTHOR)<i> eeprzhang@mail.scut.edu.cn</i><br /><searchLink fieldCode="AR" term="%22Zhang%2C+Jiaxin%22">Zhang, Jiaxin</searchLink><relatesTo>1</relatesTo> (AUTHOR)<i> msjxzhang@mail.scut.edu.cn</i><br /><searchLink fieldCode="AR" term="%22Cao%2C+Jiahuan%22">Cao, Jiahuan</searchLink><relatesTo>1</relatesTo> (AUTHOR)<i> eejiahuancao@mail.scut.edu.cn</i><br /><searchLink fieldCode="AR" term="%22Li%2C+Hongliang%22">Li, Hongliang</searchLink><relatesTo>1</relatesTo> (AUTHOR)<i> eehongliangli@mail.scut.edu.cn</i><br /><searchLink fieldCode="AR" term="%22Jin%2C+Lianwen%22">Jin, Lianwen</searchLink><relatesTo>1</relatesTo> (AUTHOR)<i> eelwjin@scut.edu.cn</i> – Name: TitleSource Label: Source Group: Src Data: <searchLink fieldCode="JN" term="%22International+Journal+of+Computer+Vision%22">International Journal of Computer Vision</searchLink>. Jul2025, Vol. 133 Issue 7, p3891-3917. 27p. – Name: Subject Label: Subjects Group: Su Data: <searchLink fieldCode="DE" term="%22Language+models%22">Language models</searchLink><br /><searchLink fieldCode="DE" term="%22Complex+variables%22">Complex variables</searchLink><br /><searchLink fieldCode="DE" term="%22Encoding%22">Encoding</searchLink> – Name: Abstract Label: Abstract Group: Ab Data: We propose LGGPT, an LLM-based model tailored for unified layout generation. First, we propose Arbitrary Layout Instruction (ALI) and Universal Layout Response (ULR) as the uniform I/O template. ALI accommodates arbitrary layout generation task inputs across multiple layout domains, enabling LGGPT to unify both task-generic and domain-generic layout generation hitherto unexplored. Collectively, ALI and ULR boast a succinct structure that forgoes superfluous tokens typically found in existing HTML-based formats, facilitating efficient instruction tuning and boosting unified generation performance. In addition, we propose an Interval Quantization Encoding (IQE) strategy that compresses ALI into a more condensed structure. IQE precisely preserves valid layout clues while eliminating the less informative placeholders, facilitating LGGPT to capture complex and variable layout generation conditions during the unified training process. Experimental results demonstrate that LGGPT achieves superior or on par performance compared to existing methods. Notably, LGGPT strikes a prominent balance between proficiency and efficiency with a compact 1.5B parameter LLM, which beats prior 7B or 175B models even in the most extensive and challenging unified scenario. Furthermore, we underscore the necessity of employing LLMs for unified layout generation and suggest that 1.5B could be an optimal parameter size by comparing LLMs of varying scales. Code is available at https://github.com/NiceRingNode/LGGPT. [ABSTRACT FROM AUTHOR] – Name: AbstractSuppliedCopyright Label: Group: Ab Data: <i>Copyright of International Journal of Computer Vision is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract.</i> (Copyright applies to all Abstracts.) |
| PLink | https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=egs&AN=185781607 |
| RecordInfo | BibRecord: BibEntity: Identifiers: – Type: doi Value: 10.1007/s11263-025-02353-2 Languages: – Code: eng Text: English PhysicalDescription: Pagination: PageCount: 27 StartPage: 3891 Subjects: – SubjectFull: Language models Type: general – SubjectFull: Complex variables Type: general – SubjectFull: Encoding Type: general Titles: – TitleFull: Smaller But Better: Unifying Layout Generation with Smaller Large Language Models. Type: main BibRelationships: HasContributorRelationships: – PersonEntity: Name: NameFull: Zhang, Peirong – PersonEntity: Name: NameFull: Zhang, Jiaxin – PersonEntity: Name: NameFull: Cao, Jiahuan – PersonEntity: Name: NameFull: Li, Hongliang – PersonEntity: Name: NameFull: Jin, Lianwen IsPartOfRelationships: – BibEntity: Dates: – D: 01 M: 07 Text: Jul2025 Type: published Y: 2025 Identifiers: – Type: issn-print Value: 09205691 Numbering: – Type: volume Value: 133 – Type: issue Value: 7 Titles: – TitleFull: International Journal of Computer Vision Type: main |
| ResultId | 1 |
