通过不确定性监控实现基于深度学习的系统的故障安全执行
Fail-Safe Execution of Deep Learning based Systems through Uncertainty Monitoring
摘要：
现代软件系统在处理复杂的非结构化输入（如图像、视频、自然语言文本或音频信号）时依赖于深度神经网络（DNN）。假设这样的输入空间非常大，学习算法的内在限制，以及某些输入的预期预测的模糊性，不仅不能保证DNN的预测总是正确的，而且开发人员必须安全地假设一个低的，但不可忽略的错误概率。基于故障安全深度学习的系统（DLS）是一种配备有通过监督程序来处理DNN故障的系统，该监督程序能够识别不应该被信任的预测，并且应该激活使DLS处于安全状态的修复过程。在本文中，我们提出了一种使用DNN不确定性估计器来实现这样一个监督器的方法。我们首先讨论了现有方法来衡量DNN的不确定性的优点和缺点，并提出了新的指标，依赖于这种方法的监督员的经验评估。然后，我们描述了我们的公开可用的工具UNCERTAINTY-WIZARD，它允许透明的不确定性估计定期tf.keras DNN。最后，我们讨论了一个大规模的研究进行了四个不同的主题，以经验验证的方法，报告的经验教训，指导软件工程师谁打算监控不确定性的故障安全执行DLS。
英文摘要：
Modern software systems rely on Deep Neural Networks (DNN) when processing complex, unstructured inputs, such as images, videos, natural language texts or audio signals. Provided the intractably large size of such input spaces, the intrinsic limitations of learning algorithms, and the ambiguity about the expected predictions for some of the inputs, not only there is no guarantee that DNN's predictions are always correct, but rather developers must safely assume a low, though not negligible, error probability. A fail-safe Deep Learning based System (DLS) is one equipped to handle DNN faults by means of a supervisor, capable of recognizing predictions that should not be trusted and that should activate a healing procedure bringing the DLS to a safe state. In this paper, we propose an approach to use DNN uncertainty estimators to implement such a supervisor. We first discuss the advantages and disadvantages of existing approaches to measure uncertainty for DNNs and propose novel metrics for the empirical assessment of the supervisor that rely on such approaches. We then describe our publicly available tool UNCERTAINTY-WIZARD, which allows transparent estimation of uncertainty for regular tf.keras DNNs. Lastly, we discuss a large-scale study conducted on four different subjects to empirically validate the approach, reporting the lessons-learned as guidance for software engineers who intend to monitor uncertainty for fail-safe execution of DLS.
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。
