Understanding The Robustness in Vision Transformers
share
Recent studies show that Vision Transformers(ViTs) exhibit strong robustness against various corruptions. Although this property is partly attributed to the self-attention mechanism, there is still a lack of systematic understanding. In this paper, we examine the role of self-attention in learning robust representations. Our study is motivated by the intriguing properties of the emerging visual grouping in Vision Transformers, which indicates that self-attention may promote robustness through improved mid-level representations. We further propose a family of fully attentional networks (FANs) that strengthen this capability by incorporating an attentional channel processing design. We validate the design comprehensively on various hierarchical backbones. Our model achieves a state of-the-art 87.1 and 35.8 demonstrate state-of-the-art accuracy and robustness in two downstream tasks: semantic segmentation and object detection. Code will be available at https://github.com/NVlabs/FAN.
READ FULL TEXT