Make up & hair
By altering the expected dark and light areas of a face (or object) according to the vulnerabilities of a specific computer vision algorithm. In the image above, the design targets the Viola-Jones face detection algorithm, a popular and open source face detector that is included with the OpenCV computer vision framework.
CV Dazzle aims to introduce low-cost methods for breaking computer vision algorithms. Here, the design are created using only hair styling, makeup, and fashion accessories, which could be customized to any wearer’s style.
Because computer vision is always based on probabilistic thresholds, these can be exploited by altering the key visual features to appear one step below the threshold of detection.
And since face detection is the first step in any automated facial recognition system, blocking the detection stage also blocks any subsequent facial analysis including recognition and emotional analysis. Therefore, CV Dazzle could be used to block facial recognition by blocking face detection using only hair styling and makeup.