comix
Datasets
All datasets matching “comix”comix-v0_1-pages
CoMix v0.1 - Pages Dataset
This is the Full CoMix dataset for page-level work. Download comix-v0_1-pages-tiny for fast experiments.
Some numbers: 19063 books, 894633 single pages, 6M+ single panels. v0.1 has a few broken tars, total number of books should be >20k).
Note: Dataset viewer currently struggles with this dataset because seg.npz files are custom NumPy archives with variable keys/shapes per page.
Will improve in following versions.
... add here an [image of the CoMix… See the full description on the dataset page: https://huggingface.co/datasets/emanuelevivoli/comix-v0_1-pages.comix-v0_1-pages-tiny
Comic Books Dataset v0.1 - Pages
Full dataset of comic book pages from Digital Comic Museum.
This is the PRODUCTION dataset. For testing, use comix_v0_tiny_pages.
What's Included
Each page has:
{page_id}.jpg - Page image
{page_id}.json - Metadata (detections, captions, page class)
{page_id}.seg.npz - Segmentation masks (SAMv2) [when available]
Quick Start
from datasets import load_dataset
# Load pages dataset
pages = load_dataset(… See the full description on the dataset page: https://huggingface.co/datasets/emanuelevivoli/comix-v0_1-pages-tiny.comix_v0_tiny_pages
Comic Books Tiny Dataset v0 - Pages (Testing)
Small test dataset of comic book pages for rapid development and testing.
⚠️ This is a TINY dataset for testing only. For production, use comix_v0_pages.
What's Included
Each page has:
{page_id}.jpg - Page image
{page_id}.json - Metadata (detections, captions, page class)
{page_id}.seg.npz - Segmentation masks (SAMv2)
Quick Start
from datasets import load_dataset
import numpy as np
# Load tiny pages dataset
pages… See the full description on the dataset page: https://huggingface.co/datasets/emanuelevivoli/comix_v0_tiny_pages.comix-v0_1-books-tiny
Comic Books Dataset v0.1 - Books
Full dataset of book-level metadata from Digital Comic Museum.
This is the PRODUCTION dataset. For testing, use comix_v0_tiny_books.
What's Included
Each book has:
{book_id}.json - Book metadata with page references
Purpose
This dataset provides book-level metadata to group pages from comix-v0_1-pages.
Workflow:
Download comix-v0_1-pages (with images)
Download comix-v0_1-books (metadata only)
Use WebDataset pipeline to group… See the full description on the dataset page: https://huggingface.co/datasets/emanuelevivoli/comix-v0_1-books-tiny.Niji_comix_stylecomix_books_v0
Comic Books Dataset v0
A large-scale multi-collection dataset of annotated comic book pages with panel detections, character bounding boxes, captions, and segmentation masks.
Collections
This dataset provides two collections for different use cases:
1. Pages Collection (pages/)
Individual page-level data for training vision models.
Files per page:
{book_id}_{page_id}.json - Metadata (detections, captions, page class)
{book_id}_{page_id}.jpg - Page image… See the full description on the dataset page: https://huggingface.co/datasets/emanuelevivoli/comix_books_v0.