研究人员开发了 Super-DeepG，一种用于在图像数据集中形式化验证神经网络免受几何扰动的新颖方法。该方法增强了线性松弛和 Lipschitz 优化等推理技术，在鲁棒性认证方面提供了更高的精度和计算效率。Super-DeepG 可在 GitHub 上作为开源工具使用，旨在确保在安全关键型应用中的预期性能。 AI
影响 增强了安全关键型人工智能应用的鲁棒性认证，提高了针对图像扰动的可靠性。
排序理由 关于神经网络形式化验证新方法的学术论文。
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
完整方法见我们的编辑标准。
arXiv:2604.24379v1 Announce Type: cross Abstract: Safety-critical applications are required to perform as expected in normal operations. Image processing functions are often required to be insensitive to small geometric perturbations such as rotation, scaling, shearing or transla…
Safety-critical applications are required to perform as expected in normal operations. Image processing functions are often required to be insensitive to small geometric perturbations such as rotation, scaling, shearing or translation. This paper addresses the formal verification…