Datasets:
metadata
annotations_creators:
- found
language:
- en
license:
- cc0-1.0
multilinguality:
- monolingual
pretty_name: piczel.tv
size_categories:
- 10K<n<100K
task_categories:
- image-classification
- text-to-image
tags:
- not-for-all-audiences
- art
- image
- gallery
Dataset Card for piczel.tv
Dataset Summary
This dataset contains information about image files from piczel.tv, an art streaming and gallery service. The dataset includes both metadata and the original media files themselves. The metadata covers 76,607 image entries, including URLs, tags, file information, upload dates and descriptions. The dataset also contains the corresponding original image and video files.
Languages
The dataset metadata is in English.
Dataset Structure
Data Fields
This dataset includes the following fields for each image file, stored in the piczel.jsonl file:
page_url: URL of the post on piczel.tv (string)image_url: Direct URL to the image file (string)tags: List of tags associated with the image (list of strings)name: Title of the image post (string)uploader: Username of the content uploader (string)upload_date: Date when the content was uploaded (string)upload_time: Time when the content was uploaded (string)description_title: Title in the description section (string)description: Full description text for the post (string)
Data Splits
All examples are in a single split.