Increasing the Trustworthiness of Deep Neural Networks via Accuracy Monitoring
Increasing the Trustworthiness of Deep Neural Networks via Accuracy Monitoring
复制标题
DOI：
--
复制
发表时间：
2020-07
期刊：
影响因子：
--
通讯作者：
Zhihui Shao;Jianyi Yang;Shaolei Ren
中科院分区：
文献类型：
--
作者：
Zhihui Shao;Jianyi Yang;Shaolei Ren
深度神经网络（dnn）的推理精度是一个关键的性能指标，但在实践中，由于实际测试数据集的不同，可能会有很大的变化，并且由于缺乏基础真值标签，通常是未知的。这引起了对深度神经网络可信度的重大关注，特别是在安全关键应用中。在本文中，我们通过使用事后处理来监控用户数据集上的真实推理准确性来解决dnn的可信度问题。具体而言，我们提出了一种基于神经网络的准确率监测模型，该模型仅以部署的DNN的softmax概率输出作为输入，直接预测DNN的预测结果是否正确，从而对真实的推理准确率进行估计。准确度监控模型可以在与目标应用相关的数据集上进行预训练，并且只需要主动标记用户数据集的一小部分（在我们的实验中为1%）进行模型转移。为了估计稳健性，我们进一步采用了基于蒙特卡罗dropout方法的监测模型集合。我们在不同部署的DNN模型上评估了我们的方法，用于多个数据集（包括对抗性样本）的图像分类和交通标志检测。结果表明，我们的精度监测模型提供了一个接近真实的精度估计，并优于现有的基线方法。
Inference accuracy of deep neural networks (DNNs) is a crucial performance metric, but can vary greatly in practice subject to actual test datasets and is typically unknown due to the lack of ground truth labels. This has raised significant concerns with trustworthiness of DNNs, especially in safety-critical applications. In this paper, we address trustworthiness of DNNs by using post-hoc processing to monitor the true inference accuracy on a user's dataset. Concretely, we propose a neural network-based accuracy monitor model, which only takes the deployed DNN's softmax probability output as its input and directly predicts if the DNN's prediction result is correct or not, thus leading to an estimate of the true inference accuracy. The accuracy monitor model can be pre-trained on a dataset relevant to the target application of interest, and only needs to actively label a small portion (1% in our experiments) of the user's dataset for model transfer. For estimation robustness, we further employ an ensemble of monitor models based on the Monte-Carlo dropout method. We evaluate our approach on different deployed DNN models for image classification and traffic sign detection over multiple datasets (including adversarial samples). The result shows that our accuracy monitor model provides a close-to-true accuracy estimation and outperforms the existing baseline methods.