UltraHR-100K: A Large-Scale High-Quality Dataset for Ultra-High-Resolution Image Synthesis

UltraHR-100K is a large-scale, high-quality dataset introduced to address the absence of such data for ultra-high-resolution (UHR) text-to-image generation. It was created by author zhihefang and was last updated on the Hugging Face platform in November 2025. The dataset is intended to support progress in fine-grained detail synthesis for UHR scenarios.

Use Cases

• Training text-to-image models for ultra-high-resolution generation based on the described high-quality image-text pairs.
• Developing tailored training strategies for fine-grained detail synthesis based on the dataset's UHR image content.
• Benchmarking the performance of image synthesis models on high-resolution outputs.
• Researching the relationship between dataset scale and quality in UHR image generation.

Strengths

• Dataset is explicitly described as 'large-scale' and 'high-quality' for its domain.
• It directly addresses a noted challenge in the field: the absence of a large-scale UHR T2I dataset.
• The dataset was last updated on 2025-11-21, indicating recent maintenance.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The full description is hosted externally, requiring a click-through for complete details.

Provenance

Source

Hugging Face, uploaded by author zhihefang.

Collection Method

Method of collection is not specified in the provided input.

Freshness

Last updated 2025-11-21 07:26:22.