Anime Inpainting Data: 9,734 Images for Tiled Reference Completion

LAXMAYDAY's Anime Inpainting Data contains 9,734 reproducibly sampled anime images for reference-based image completion tasks. Each sample includes a deterministic partial crop of a source image, an English instruction prompt, and the complete target image. The dataset was generated with seed 251100 and was last updated on June 15, 2026.

Use Cases

• Train image inpainting models based on partial image crops and text prompts.
• Benchmark reference-based image completion algorithms using deterministic crops.
• Develop conditional generative models for anime art based on textual instructions.
• Study the relationship between partial visual context and full image generation.

Strengths

• Contains 9,734 anime images, providing a substantial collection for model training.
• Data generation is reproducible, using a fixed seed (251100) for sampling.
• Each sample includes a text prompt, enabling multimodal (text-to-image) training.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Last updated 2026-06-15 09:16:04; freshness should be verified.

Provenance

Source

LAXMAYDAY on Hugging Face.

Collection Method

Reproducibly sampled anime images, generated with a specific seed.

Freshness

Last updated 2026-06-15 09:16:04.