Datasets

A comprehensive dataset for testing anime-style base models' character generation capabilities through extracted features and metadata from 7,371 diverse characters.

A large-scale anime image dataset with 11.4 million WebP-encoded images and pre-computed visual embeddings for computer vision and generative AI applications.

A comprehensive collection of character skin and costume artwork from multiple popular mobile games for computer vision and image processing research.

This repository provides index files for downloading and accessing the Danbooru2023 WebP 4Mpixel dataset containing millions of anime-style images.

A comprehensive dataset containing 23 million Midjourney-generated images paired with their text prompts for multimodal AI research.

This repository provides a comprehensive dataset of anime and illustration images with detailed metadata for computer vision and machine learning applications.

This dataset provides 8.05 million anime-style images with comprehensive metadata for training and evaluating computer vision models in multi-modal learning applications.

This dataset provides image embeddings and prediction results for training models to invert tag predictions back to embedding space for semantic image search applications.

A comprehensive dataset integrating anime metadata from SubsPlease, MyAnimeList, and Nyaa.si for tracking current anime trends and availability.

A comprehensive collection of structured tag metadata from 18 popular image hosting platforms for multilingual tag analysis and content classification research.

A comprehensive dataset containing character information, images, and tags from popular mobile games for character recognition and image classification tasks.

This is a comprehensive dataset containing 11.4 million anime-style images from Gelbooru with extensive tag annotations for machine learning research.

A comprehensive collection of anime character images from multiple franchises sourced from Zerochan for character recognition and art analysis tasks.

This repository provides index files for the Danbooru2023 dataset containing metadata for millions of anime-style images with detailed tag annotations.

This dataset provides 8.05 million anime and illustration images with comprehensive metadata and tags for computer vision and multi-modal AI research.

A comprehensive dataset containing over 11 million explicit anime-style images with detailed metadata and tag annotations from rule34.xxx.

A large-scale anime image dataset containing 16+ million WebP-encoded images with visual embeddings for computer vision research.

A comprehensive dataset of 4.1 million anime-style images with detailed character, series, and thematic annotations for computer vision research.

This dataset provides a comprehensive collection of 29+ million anime images with detailed metadata for computer vision and anime content analysis.

A large-scale anime image dataset containing 4.1 million WebP-encoded images with pre-computed embeddings for computer vision research and content generation.

This dataset contains over 22 million anime-style images with comprehensive metadata for training image classification and generation models.

This repository contains a comprehensive dataset of anime-style images with extensive metadata and tagging for multi-modal AI research and content classification.

A large-scale dataset of anime-style images in WebP format with pre-computed embeddings for computer vision and generative AI applications.

A large-scale anime-style image dataset with 11+ million WebP-encoded images and pre-computed embeddings for computer vision and generative AI applications.

A large-scale anime image dataset with WebP-encoded images and pre-computed embeddings for computer vision research.

This repository provides index files for efficiently accessing and downloading the e621-2024-webp-4Mpixel image dataset through the cheesechaser library.

A large-scale anime image dataset containing over 22 million WebP-encoded images with pre-computed embeddings for machine learning applications.

A curated dataset of 6.5 million anime and illustration images with comprehensive metadata for image classification and text-to-image generation tasks.

A comprehensive dataset containing the full collection of anime-style images from Safebooru.org with extensive metadata and tag annotations.

A comprehensive dataset of 3.8 million human-centric images from Reddit with detailed metadata for computer vision research and model training.

A comprehensive dataset integrating MyAnimeList metadata with Fancaps screenshot collections for multi-modal anime analysis and research.

This dataset provides continuously updated anthropomorphic artwork and multimedia content from e621.net with comprehensive metadata for computer vision research.

A curated dataset of anime-style images optimized for computer vision research and generative AI applications.

A dataset for training models to determine whether two anime images containing single characters depict the same character.

A comprehensive dataset of 7.9 million popular DeviantArt images with captions and metadata for computer vision and multimodal AI research.

A large-scale anime image dataset with 5.7 million WebP-encoded images and SwinV2 embeddings for computer vision and generative AI applications.

A comprehensive dataset of 642,980 anime-style images with detailed metadata for computer vision and image generation research.

A curated dataset of anime background and wallpaper images organized for efficient access and use in image generation applications.

A large-scale anime image dataset with 29 million WebP-encoded images and pre-computed embeddings for computer vision applications.

This dataset provides a comprehensive collection of anime-style images with detailed metadata and tag annotations for multi-modal AI research and content generation.

A multi-class image classification dataset designed for training NSFW content detection models with fine-grained categorization of artistic and sensitive content.

A curated dataset for training monochrome image classification models using artwork from danbooru and pixiv.

A comprehensive anime metadata dataset integrating information from Erai-raws and MyAnimeList with cross-platform identifiers and publication details.

This dataset provides character frequency statistics for 116 languages derived from the CC-100 corpus for font support analysis and language processing applications.

A curated dataset of 319,012 anime-style images in WebP format with pre-computed embeddings for computer vision applications.

This repository provides index files for the e621-2024 dataset, enabling efficient downloading and management of millions of images for computer vision research.

This dataset provides a comprehensive collection of anime-style images with detailed metadata and tagging for text-to-image generation and image classification research.

This dataset provides a comprehensive collection of anthropomorphic and furry-themed multimedia content with detailed metadata annotations for AI research and content analysis.

A comprehensive metadata collection for anime image tagging systems containing multi-language tag translations and detailed attribute annotations.

This repository provides index metadata for efficient selective access to individual images within the yandere2023 dataset archives.

A comprehensive dataset containing the full wiki documentation and tag metadata from Danbooru for training language models and tag classification systems.

This repository contains test assets for unit testing the soundutils audio processing library.

A Japanese voice-text dataset containing 30,800 audio samples from Fate/Grand Order characters for automatic speech recognition and voice analysis applications.

A Japanese voice-text dataset containing 10,905 audio recordings from Arknights game characters for automatic speech recognition and voice synthesis applications.

A comprehensive Chinese voice-text dataset featuring character dialogues from the Arknights game for speech recognition and audio classification tasks.

A specialized dataset for anime face detection containing annotated images with bounding boxes for training object detection models.

A comprehensive dataset for anime head detection containing multiple versions with high-quality manual annotations for training object detection models.

A Japanese voice-text dataset containing character dialogues from Girls Front Line for automatic speech recognition and audio classification tasks.

A Japanese voice-text dataset containing 30,160 audio samples from Azur Lane characters for automatic speech recognition and audio classification tasks.

An English voice-text dataset containing character dialogues from the Arknights mobile game for automatic speech recognition and audio classification tasks.

A specialized dataset for detecting anime characters in various artistic styles including illustrations, comics, and anime series.

This dataset provides labeled images for binary classification of monochrome versus colored images.

A comprehensive collection of anime-style images designed for style classification, artist recognition, and contrastive learning applications in computer vision.

A cleaned anime style image dataset designed for zero-shot classification and contrastive learning tasks in computer vision.

A Korean voice-text dataset extracted from Arknights game characters for automatic speech recognition and voice classification tasks.

A large-scale dataset for robust complex anime scene text detection containing multilingual text annotations and hard negative samples.

This repository provides structured vocabulary datasets for image tagging models to support natural language processing and computer vision applications.

This dataset integrates anime metadata with torrent availability data to track popular anime series and their download accessibility.

A dataset for detecting censorship elements in anime-style artwork using object detection models.

This repository provides a human-curated evaluation dataset of anime-style images for benchmarking zero-shot image classification models on artistic style recognition.

A specialized dataset for training object detection models to identify and localize anime character halfbodies in images.

A large-scale dataset for anime image tagging and classification with comprehensive metadata annotations.

This dataset provides a large-scale collection of anthropomorphic artwork with comprehensive metadata annotations for multi-modal AI research.

This dataset provides categorized tag metadata from Sankaku Complex to enhance WD14Tagger's image annotation capabilities for anime-style artwork.

This is a binary classification dataset for distinguishing between clear and blurry 3D anime-style images using automatic BLIP annotation.

A large-scale dataset for anime image rating classification using the danbooru tagging system to train content moderation and image classification models.

This dataset provides annotated bounding boxes for detecting eyes in anime-style character images, formatted for object detection models.

This repository provides a curated index of timm models organized by size levels for efficient computer vision model selection.