Investigate Other Redaction Libraries

[KrishPatel13]

Feature request

Presidio Image Redaction is not completely reliable right now. As it not redacting all PII/PHI.

Motivation

I am frustrated when I investigated https://github.com/MLDSAI/OpenAdapt/pull/328#issuecomment-1611909662 so this feature would Investigate different Image Redaction Libraries, testing on sample Images where Presidio Image Redaction Fails:

eg. https://github.com/MLDSAI/OpenAdapt/pull/328#issuecomment-1611909662

[KrishPatel13]

1. OpenRedact
   • Github: https://github.com/openredact/anonymizer
   • Article: https://medium.com/@openredact/anonymizer-a-framework-for-text-anonymization-499855f639d4
   • Other: Related to #272
     • From Article:

[KrishPatel13]

2. Open-Data-Anonymizer
   • Github: https://github.com/ArtLabss/open-data-anonymizer
   • Useful for Redaction of People Faces (personal images), PDF.
   • See some nice usage examples:
     • https://github.com/ArtLabss/open-data-anonymizer

Some Demo Pictures: PDF Redaction:

Personal Image:

[KrishPatel13]

3. BMW-Anonymization-API
   • Github: https://github.com/BMW-InnovationLab/BMW-Anonymization-API
   • Useful for Video Redaction (specifically, object anonymization like. People faces, identity. etc. )
     • https://github.com/BMW-InnovationLab/BMW-Anonymization-API/blob/master/references/output_7.gif

[KrishPatel13]

Cape-Privacy is good!

• Website: https://docs.capeprivacy.com/getting-started

• API Reference: https://api.capeprivacy.com/v1/redoc

• "The API exposes granular control of essential privacy tools including de-identification and re-identification, supporting redaction for 50 types of entities like names, emails, credit cards, addresses, and other PII/PCI/PHI across 47 languages." claims Cape-Privacy

• Examples:

  

[abrichr]

https://github.com/OpenAdaptAI/OpenAdapt/pull/418

[KrishPatel13]

After #486 shall we close this ?