Anime Character Eye Bounding Box Annotations

An ultralytics-compatible dataset provides annotated bounding boxes for eyes in anime-style character images. The dataset is formatted for YOLOv8 and other Ultralytics frameworks, enabling training of computer vision models for eye detection in anime artwork. It was created by author deepghs and last updated in November 2025.

Use Cases

• Train a YOLOv8 model on bounding box annotations to detect and localize eyes in anime character images.
• Fine-tune an object detection model for anime-specific eye features using the provided ultralytics-compatible format.
• Analyze the distribution and scale of eye bounding boxes across a collection of anime artwork.
• Build a preprocessing pipeline for anime portraits that crops and aligns images based on detected eye positions.

Strengths

• Dataset is specifically formatted for seamless use with YOLOv8 and Ultralytics frameworks.
• Annotations are focused on a distinct domain: anime-style character eyes.
• The dataset was updated in November 2025, indicating recent maintenance.

Limitations

• The dataset size is categorized as 'n1 K', suggesting it contains under 1,000 images, which is a small sample for training robust models.
• Specifics on annotation quality, label consistency, or image source diversity are not provided.
• No information is available on the number of bounding boxes per image or class balance.

Provenance

Source

Hugging Face dataset repository by author deepghs.

Collection Method

null

Time Range

null

Freshness

Last updated on 2025-11-17.

Geography

null