隐私保护可视化问答
Privacy Preserving Visual Question Answering
摘要:
我们引入了一种新的隐私保护方法,用于在边缘执行视觉问答。我们的方法构建了一个符号表示的视觉场景,使用低复杂度的计算机视觉模型,共同预测类,属性和谓词。这种符号表示是不可微的,这意味着它不能用于恢复原始图像,从而保持原始图像的私密性。我们提出的混合解决方案使用的视觉模型比当前最先进的(SOTA)视觉模型小25倍以上,比端到端SOTA VQA模型小100倍。我们报告详细的错误分析,并讨论了使用蒸馏视觉模型和视觉场景的符号表示的权衡。
英文摘要:
We introduce a novel privacy-preserving methodology for performing Visual Question Answering on the edge. Our method constructs a symbolic representation of the visual scene, using a low-complexity computer vision model that jointly predicts classes, attributes and predicates. This symbolic representation is non-differentiable, which means it cannot be used to recover the original image, thereby keeping the original image private. Our proposed hybrid solution uses a vision model which is more than 25 times smaller than the current state-of-the-art (SOTA) vision models, and 100 times smaller than end-to-end SOTA VQA models. We report detailed error analysis and discuss the trade-offs of using a distilled vision model and a symbolic representation of the visual scene.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。