UltraHR-100K: Enhancing UHR Image Synthesis with A Large-Scale High-Quality Dataset
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866915572154892288 |
|---|---|
| author | Zhao, Chen Ci, En Xu, Yunzhe Fan, Tiehan Guan, Shanyan Ge, Yanhao Yang, Jian Tai, Ying |
| author_facet | Zhao, Chen Ci, En Xu, Yunzhe Fan, Tiehan Guan, Shanyan Ge, Yanhao Yang, Jian Tai, Ying |
| contents | Ultra-high-resolution (UHR) text-to-image (T2I) generation has seen notable progress. However, two key challenges remain : 1) the absence of a large-scale high-quality UHR T2I dataset, and (2) the neglect of tailored training strategies for fine-grained detail synthesis in UHR scenarios. To tackle the first challenge, we introduce \textbf{UltraHR-100K}, a high-quality dataset of 100K UHR images with rich captions, offering diverse content and strong visual fidelity. Each image exceeds 3K resolution and is rigorously curated based on detail richness, content complexity, and aesthetic quality. To tackle the second challenge, we propose a frequency-aware post-training method that enhances fine-detail generation in T2I diffusion models. Specifically, we design (i) \textit{Detail-Oriented Timestep Sampling (DOTS)} to focus learning on detail-critical denoising steps, and (ii) \textit{Soft-Weighting Frequency Regularization (SWFR)}, which leverages Discrete Fourier Transform (DFT) to softly constrain frequency components, encouraging high-frequency detail preservation. Extensive experiments on our proposed UltraHR-eval4K benchmarks demonstrate that our approach significantly improves the fine-grained detail quality and overall fidelity of UHR image generation. The code is available at \href{https://github.com/NJU-PCALab/UltraHR-100k}{here}. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_20661 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | UltraHR-100K: Enhancing UHR Image Synthesis with A Large-Scale High-Quality Dataset Zhao, Chen Ci, En Xu, Yunzhe Fan, Tiehan Guan, Shanyan Ge, Yanhao Yang, Jian Tai, Ying Computer Vision and Pattern Recognition Ultra-high-resolution (UHR) text-to-image (T2I) generation has seen notable progress. However, two key challenges remain : 1) the absence of a large-scale high-quality UHR T2I dataset, and (2) the neglect of tailored training strategies for fine-grained detail synthesis in UHR scenarios. To tackle the first challenge, we introduce \textbf{UltraHR-100K}, a high-quality dataset of 100K UHR images with rich captions, offering diverse content and strong visual fidelity. Each image exceeds 3K resolution and is rigorously curated based on detail richness, content complexity, and aesthetic quality. To tackle the second challenge, we propose a frequency-aware post-training method that enhances fine-detail generation in T2I diffusion models. Specifically, we design (i) \textit{Detail-Oriented Timestep Sampling (DOTS)} to focus learning on detail-critical denoising steps, and (ii) \textit{Soft-Weighting Frequency Regularization (SWFR)}, which leverages Discrete Fourier Transform (DFT) to softly constrain frequency components, encouraging high-frequency detail preservation. Extensive experiments on our proposed UltraHR-eval4K benchmarks demonstrate that our approach significantly improves the fine-grained detail quality and overall fidelity of UHR image generation. The code is available at \href{https://github.com/NJU-PCALab/UltraHR-100k}{here}. |
| title | UltraHR-100K: Enhancing UHR Image Synthesis with A Large-Scale High-Quality Dataset |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2510.20661 |