Egy-DRiVeS uses Roop face-swapping to preserve pedestrian privacy and data usability

This work implements a five-stage pipeline for the Egy-DRiVeS dataset that protects pedestrian privacy through face swapping while maintaining essential facial attributes required for training autonomous vehicle models.

The study evaluates Roop and Ghost-v2 face-swapping models to balance identity concealment with data usability. The analysis proves that Roop outperforms Ghost-v2 in various aspects, making it the selected model for the pipeline.

This approach addresses the challenge of applying privacy preservation procedures without degrading image usability, which hinders the effectiveness of pedestrian intention and trajectory prediction models.