Skip to content

Dataset · hyperbots-inc

Synthetic PII redaction documents dataset

hyperbots-inc/synthetic-pii-redaction

Corporate document pages with sensitive data markup for tokenization and detection.

What's inside

Contains fewer than 1K images (100 pages) of English corporate documents (KYC forms, medical records, insurance claims, bank statements). Each PII value is annotated twice: as a text range and as pixel bounding boxes on the image.

What you can do with it

  • Training token-classification models for entity extraction
  • Training object-detection models for sensitive data search
  • Developing automated personal data masking systems
Tasks
Token classification, Object detection
Languages
English
Size
up to 1K rows
Downloads this month
504
Likes
46

How to load it

from datasets import load_dataset

ds = load_dataset("hyperbots-inc/synthetic-pii-redaction")