FRPSS: Feature Rearrangement in Pre-Shape Space for Single-Image Generation
Not on the map. Why?
Abstract (arXiv)
Generative models trained on a single image often struggle to balance global structural integrity and local diversity. Existing single-image generation methods commonly rely on random noise to drive the generation process and lack explicit global structural constraints, making the generated results prone to spatial structural misalignment when structural variations occur. To address the issue, Feature Rearrangement in Pre-Shape Space for Single-Image Generation (FRPSS) is proposed in this paper. The core of FRPSS is the Manifold Structural Rearrangement with Feature Augmentation on Geodesic Surface (MSR-FAGS) module. MSR-FAGS replaces the randomly initialized features of the low-scale generator with rearranged Pre-Shape features and uses the features to guide image generation at subsequent scales, thereby reducing the risk of structural misalignment. To support downstream tasks such as stylization, a Scale-adaptive Sliding-window Patch Extraction (SSPE) strategy is further designed, and a directional Contrastive Language-Image Pre-training supervision module with SSPE (CLIP-SSPE) is constructed. Qualitative and quantitative experiments demonstrate that FRPSS achieves the best Single Image Fr'echet Inception Distance (SIFID) scores on all three datasets while maintaining competitive Learned Perceptual Image Patch Similarity (LPIPS). Further qualitative experiments verify the effectiveness of FRPSS across multiple downstream tasks with the CLIP-SSPE module.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 27 of its 42 references that are in the map. First posted on arXiv 15 Sep 2026.
Builds on
27 of 42 references.
References · 27
- 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
- 2020Denoising Diffusion Probabilistic Modelscited by 21,076
- 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,177
- 2015Rethinking the Inception Architecture for Computer Visioncited by 5,475
- 2019Analyzing and Improving the Image Quality of StyleGANcited by 3,588
- 2017Improved Training of Wasserstein GANscited by 2,235
- 2021StyleCLIP: Text-Driven Manipulation of StyleGAN Imagerycited by 991
- 2021DiffusionCLIP: Text-Guided Diffusion Models for Robust Image Manipulationcited by 642
- 2019SinGAN: Learning a Generative Model from a Single Natural Imagecited by 348
- 2022Splicing ViT Features for Semantic Appearance Transfercited by 209
- 2021CLIPstyler: Image Style Transfer with a Single Text Conditioncited by 203
- 2020Improved Techniques for Training Single-Image GANscited by 78
15 more
- 2022SinDDM: A Single Image Denoising Diffusion Modelcited by 60
- 2022SinFusion: Training Diffusion Models on a Single Image or Videocited by 58
- 2024InstantStyle-Plus: Style Transfer with Content-Preserving in Text-to-Image Generationcited by 53
- 2021Drop the GAN: In Defense of Patches Nearest Neighbors as Single Image Generative Modelscited by 51
- 2020Hierarchical Patch VAE-GAN: Generating Diverse Videos from a Single Samplecited by 43
- 2022SinDiffusion: Learning a Diffusion Model from a Single Natural Imagecited by 42
- 2022Generating natural images with direct Patch Distributions Matchingcited by 27
- 2023Image Data Augmentation Approaches: A Comprehensive Survey and Future directionscited by 23
- 2022PetsGAN: Rethinking Priors for Single Image Generationcited by 14
- 2021ExSinGAN: Learning an Explainable Generative Model from a Single Imagecited by 12
- 2021MOGAN: Morphologic-structure-aware Generative Learning from a Single Imagecited by 9
- 2023Deep Internal Learning: Deep Learning from a Single Inputcited by 9
- 2022A Kendall Shape Space Approach to 3D Shape Estimation from 2D Landmarkscited by 3
- 2026StructDiff: A Structure-Preserving and Spatially Controllable Diffusion Model for Single-Image Generationcited by 1
- 2026Efficient and Training-Free Single-Image Diffusion Modelscited by 1
Built on by
Not cited yet.