Personal data protection
Automated anonymization of people in images — detecting and masking faces and sensitive data, shared with the community on Kaggle.
Business value
_You're one step away
The project aimed to process a previously collected video database by blurring the faces of third parties and removing personal data from the recordings. This allowed the data to be published ethically and in compliance with applicable law, supporting the growth of transparent scientific research.
_You take the step
We processed the recordings using an advanced algorithm published in 2022 at the prestigious CVPR conference. The anonymization process ran on a purpose-built computer with an NVIDIA graphics card, which made it possible to process over 3 million frames in two weeks. In the future we plan to use the PL Grid infrastructure and the Cyfronet supercomputer for more complex projects.
_You've taken the step
The database we created was published on the Kaggle platform, enabling its broad use by the scientific community. This supports transparent and reproducible research in the field of artificial intelligence. Several researchers have already begun working with the published data, underscoring its usefulness and value for the advancement of science.
Faces detected
_Face detection
Crowd scene
_After anonymization
Sensitive data masked
_Anonymization process
Repeatable pipeline
