Learning Quantized Adaptive Conditions for Diffusion Models
| Title: | Learning Quantized Adaptive Conditions for Diffusion Models |
|---|---|
| Authors: | Liang, Yuchen, Tian, Yuchuan, Yu, Lei, Tang, Huao, Hu, Jie, Fang, Xiangzhong, Chen, Hanting |
| Publication Year: | 2024 |
| Collection: | Computer Science |
| Subject Terms: | Computer Science - Computer Vision and Pattern Recognition |
| More Details: | The curvature of ODE trajectories in diffusion models hinders their ability to generate high-quality images in a few number of function evaluations (NFE). In this paper, we propose a novel and effective approach to reduce trajectory curvature by utilizing adaptive conditions. By employing a extremely light-weight quantized encoder, our method incurs only an additional 1% of training parameters, eliminates the need for extra regularization terms, yet achieves significantly better sample quality. Our approach accelerates ODE sampling while preserving the downstream task image editing capabilities of SDE techniques. Extensive experiments verify that our method can generate high quality results under extremely limited sampling costs. With only 6 NFE, we achieve 5.14 FID on CIFAR-10, 6.91 FID on FFHQ 64x64 and 3.10 FID on AFHQv2. |
| Document Type: | Working Paper |
| Access URL: | http://arxiv.org/abs/2409.17487 |
| Accession Number: | edsarx.2409.17487 |
| Database: | arXiv |
| FullText | Text: Availability: 0 CustomLinks: – Url: http://arxiv.org/abs/2409.17487 Name: EDS - Arxiv Category: fullText Text: View this record from Arxiv MouseOverText: View this record from Arxiv – Url: https://resolver.ebsco.com/c/xy5jbn/result?sid=EBSCO:edsarx&genre=article&issn=&ISBN=&volume=&issue=&date=20240925&spage=&pages=&title=Learning Quantized Adaptive Conditions for Diffusion Models&atitle=Learning%20Quantized%20Adaptive%20Conditions%20for%20Diffusion%20Models&aulast=Liang%2C%20Yuchen&id=DOI: Name: Full Text Finder (for New FTF UI) (s8985755) Category: fullText Text: Find It @ SCU Libraries MouseOverText: Find It @ SCU Libraries |
|---|---|
| Header | DbId: edsarx DbLabel: arXiv An: edsarx.2409.17487 RelevancyScore: 1112 AccessLevel: 3 PubType: Report PubTypeId: report PreciseRelevancyScore: 1112.25854492188 |
| IllustrationInfo | |
| Items | – Name: Title Label: Title Group: Ti Data: Learning Quantized Adaptive Conditions for Diffusion Models – Name: Author Label: Authors Group: Au Data: <searchLink fieldCode="AR" term="%22Liang%2C+Yuchen%22">Liang, Yuchen</searchLink><br /><searchLink fieldCode="AR" term="%22Tian%2C+Yuchuan%22">Tian, Yuchuan</searchLink><br /><searchLink fieldCode="AR" term="%22Yu%2C+Lei%22">Yu, Lei</searchLink><br /><searchLink fieldCode="AR" term="%22Tang%2C+Huao%22">Tang, Huao</searchLink><br /><searchLink fieldCode="AR" term="%22Hu%2C+Jie%22">Hu, Jie</searchLink><br /><searchLink fieldCode="AR" term="%22Fang%2C+Xiangzhong%22">Fang, Xiangzhong</searchLink><br /><searchLink fieldCode="AR" term="%22Chen%2C+Hanting%22">Chen, Hanting</searchLink> – Name: DatePubCY Label: Publication Year Group: Date Data: 2024 – Name: Subset Label: Collection Group: HoldingsInfo Data: Computer Science – Name: Subject Label: Subject Terms Group: Su Data: <searchLink fieldCode="DE" term="%22Computer+Science+-+Computer+Vision+and+Pattern+Recognition%22">Computer Science - Computer Vision and Pattern Recognition</searchLink> – Name: Abstract Label: Description Group: Ab Data: The curvature of ODE trajectories in diffusion models hinders their ability to generate high-quality images in a few number of function evaluations (NFE). In this paper, we propose a novel and effective approach to reduce trajectory curvature by utilizing adaptive conditions. By employing a extremely light-weight quantized encoder, our method incurs only an additional 1% of training parameters, eliminates the need for extra regularization terms, yet achieves significantly better sample quality. Our approach accelerates ODE sampling while preserving the downstream task image editing capabilities of SDE techniques. Extensive experiments verify that our method can generate high quality results under extremely limited sampling costs. With only 6 NFE, we achieve 5.14 FID on CIFAR-10, 6.91 FID on FFHQ 64x64 and 3.10 FID on AFHQv2. – Name: TypeDocument Label: Document Type Group: TypDoc Data: Working Paper – Name: URL Label: Access URL Group: URL Data: <link linkTarget="URL" linkTerm="http://arxiv.org/abs/2409.17487" linkWindow="_blank">http://arxiv.org/abs/2409.17487</link> – Name: AN Label: Accession Number Group: ID Data: edsarx.2409.17487 |
| PLink | https://login.libproxy.scu.edu/login?url=https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&scope=site&db=edsarx&AN=edsarx.2409.17487 |
| RecordInfo | BibRecord: BibEntity: Subjects: – SubjectFull: Computer Science - Computer Vision and Pattern Recognition Type: general Titles: – TitleFull: Learning Quantized Adaptive Conditions for Diffusion Models Type: main BibRelationships: HasContributorRelationships: – PersonEntity: Name: NameFull: Liang, Yuchen – PersonEntity: Name: NameFull: Tian, Yuchuan – PersonEntity: Name: NameFull: Yu, Lei – PersonEntity: Name: NameFull: Tang, Huao – PersonEntity: Name: NameFull: Hu, Jie – PersonEntity: Name: NameFull: Fang, Xiangzhong – PersonEntity: Name: NameFull: Chen, Hanting IsPartOfRelationships: – BibEntity: Dates: – D: 25 M: 09 Type: published Y: 2024 |
| ResultId | 1 |