Towards Robust and Accurate Visual Prompting
This work addresses the robustness-generalization trade-off in visual prompting for vision tasks, offering an incremental improvement with practical benefits.
The paper tackles the problem of visual prompting from robust source models, finding that it inherits robustness but suffers from generalization decline, and introduces Prompt Boundary Loose (PBL) to improve standard accuracy without losing adversarial robustness.
Visual prompting, an efficient method for transfer learning, has shown its potential in vision tasks. However, previous works focus exclusively on VP from standard source models, it is still unknown how it performs under the scenario of a robust source model: Whether a visual prompt derived from a robust model can inherit the robustness while suffering from the generalization performance decline, albeit for a downstream dataset that is different from the source dataset? In this work, we get an affirmative answer of the above question and give an explanation on the visual representation level. Moreover, we introduce a novel technique named Prompt Boundary Loose (PBL) to effectively mitigates the suboptimal results of visual prompt on standard accuracy without losing (or even significantly improving) its adversarial robustness when using a robust model as source model. Extensive experiments across various datasets show that our findings are universal and demonstrate the significant benefits of our proposed method.