Attention-based Adversarial Appearance Learning of Augmented Pedestrians
This work addresses the domain adaptation challenge for pedestrian recognition in autonomous driving, but it appears incremental as it builds on existing augmentation and adversarial training techniques.
The paper tackles the sim2real domain shift problem in synthetic data for autonomous driving by proposing an attention-based adversarial method to generate realistic augmented pedestrians, resulting in improved visual realism and semantic consistency for pedestrian recognition.
Synthetic data became already an essential component of machine learning-based perception in the field of autonomous driving. Yet it still cannot replace real data completely due to the sim2real domain shift. In this work, we propose a method that leverages the advantages of the augmentation process and adversarial training to synthesize realistic data for the pedestrian recognition task. Our approach utilizes an attention mechanism driven by an adversarial loss to learn domain discrepancies and improve sim2real adaptation. Our experiments confirm that the proposed adaptation method is robust to such discrepancies and reveals both visual realism and semantic consistency. Furthermore, we evaluate our data generation pipeline on the task of pedestrian recognition and demonstrate that generated data resemble properties of the real domain.