Case Study: Valeo achieves privacy-compliant fisheye dataset anonymization with brighter AI's Deep Natural Anonymization

Valeo uses brighter AI’s Deep Natural Anonymization for first extensive automotive fisheye dataset

Valeo, a global automotive supplier, faced the challenge of creating a large-scale training dataset with fisheye camera images while strictly adhering to global privacy regulations. Traditional anonymization methods like pixelation were unsuitable as they create image artifacts that harm machine learning model quality. To resolve this, Valeo turned to brighter AI and its Deep Natural Anonymization technology to protect personal information without compromising the dataset's utility.

brighter AI's solution accurately anonymized faces and license plates by replacing them with AI-generated alternatives, working perfectly with the unique fisheye format. Deployed on-premise for full data control, the technology enabled the successful creation of the extensive WoodScape dataset, which consists of over 10,000 annotated images and is fully compatible with machine learning workflows, all while ensuring complete privacy compliance.

Valeo