Adaptive Deep Learning for Soft Real-Time Image Classification
Adaptive Deep Learning for Soft Real-Time Image Classification
复制标题
DOI：
10.3390/technologies9010020
复制
发表时间：
2021-03
期刊：
影响因子：
3.6
通讯作者：
Fangming Chai;K. Kang
中科院分区：
文献类型：
--
作者：
Fangming Chai;K. Kang
CNN（卷积神经网络）对于实时应用变得越来越重要，例如交通控制中的图像分类、视觉监控和智能制造。然而，由于 CNN 的复杂性，要满足使用 CNN 进行图像处理任务的时间限制具有挑战性。在 CNN 中进行图像数据分析的推理精度和时间之间的动态权衡也具有挑战性，因为我们通过使用两个流行的数据集 MNIST 和 CIFAR-10 评估数百个 CNN 模型的时间和精度，发现运行时间更长的更复杂的 CNN 在许多情况下甚至会导致精度更低。为了应对这些挑战，我们提出了一种新方法：(1) 生成 CNN 模型并分析其图像分类的平均推理时间和准确性，(2) 离线存储具有单调时间和准确性关系的一小部分 CNN，以及 (3) 有效地选择一个有效的 CNN，预计能够在存储的 CNN 中支持最高可能的准确性，但受运行时截止日期的剩余时间影响。在我们的广泛评估中，我们验证了通过我们的方法得出的 CNN 比两种基线方法更灵活且更具成本效益。我们验证了我们的方法可以有效地构建一组紧凑的 CNN，并在必要时有效支持系统时间与精度的权衡，以满足用户指定的时序和精度要求。此外，就延迟和内存消耗而言，我们的方法的开销很小/可以接受。
CNNs (Convolutional Neural Networks) are becoming increasingly important for real-time applications, such as image classification in traffic control, visual surveillance, and smart manufacturing. It is challenging, however, to meet timing constraints of image processing tasks using CNNs due to their complexity. Performing dynamic trade-offs between the inference accuracy and time for image data analysis in CNNs is challenging too, since we observe that more complex CNNs that take longer to run even lead to lower accuracy in many cases by evaluating hundreds of CNN models in terms of time and accuracy using two popular data sets, MNIST and CIFAR-10. To address these challenges, we propose a new approach that (1) generates CNN models and analyzes their average inference time and accuracy for image classification, (2) stores a small subset of the CNNs with monotonic time and accuracy relationships offline, and (3) efficiently selects an effective CNN expected to support the highest possible accuracy among the stored CNNs subject to the remaining time to the deadline at run time. In our extensive evaluation, we verify that the CNNs derived by our approach are more flexible and cost-efficient than two baseline approaches. We verify that our approach can effectively build a compact set of CNNs and efficiently support systematic time vs. accuracy trade-offs, if necessary, to meet the user-specified timing and accuracy requirements. Moreover, the overhead of our approach is little/acceptable in terms of latency and memory consumption.