Designing Adaptive Neural Networks for Energy-Constrained Image Classification
Designing Adaptive Neural Networks for Energy-Constrained Image Classification
复制标题
DOI：
10.1145/3240765.3240796
复制
发表时间：
2018-08
期刊：
影响因子：
--
通讯作者：
Dimitrios Stamoulis;Ting-Wu Chin;Anand P. Krishnan;Haocheng Fang;S. Sajja;Mitchell Bognar;Diana Marculescu
中科院分区：
文献类型：
--
作者：
Dimitrios Stamoulis;Ting-Wu Chin;Anand P. Krishnan;Haocheng Fang;S. Sajja;Mitchell Bognar;Diana Marculescu
随着卷积神经网络（CNNS）启用了最先进的计算机视觉应用，它们的高能量消耗已成为对其在嵌入式和移动设备上部署的关键障碍。自适应CNN，即具有不同准确性和计算特征的网络系统，其中选择方案可自适应地选择要评估每个输入图像的网络以前的努力已经调查了不同的网络选择方案，我们发现它们在移动系统上部署时不一定会节省能源。 ，每个网络都被视为黑框。被视为全球优化的超参数。对于设计空间的属性，在几十个功能评估中达到了近乎最佳的配置。 6×，与使用CNN作为Blackbox的最佳先前发布的工作，我们评估了两个图像分类实践，即在能量和通信约束下对云上的所有图像进行分类。
As convolutional neural networks (CNNs) enable state-of-the-art computer vision applications, their high energy consumption has emerged as a key impediment to their deployment on embedded and mobile devices. Towards efficient image classification under hardware constraints, prior work has proposed adaptive CNNs, i.e., systems of networks with different accuracy and computation characteristics, where a selection scheme adaptively selects the network to be evaluated for each input image. While previous efforts have investigated different network selection schemes, we find that they do not necessarily result in energy savings when deployed on mobile systems. The key limitation of existing methods is that they learn only how data should be processed among the CNNs and not the network architectures, with each network being treated as a blackbox. To address this limitation, we pursue a more powerful design paradigm where the architecture settings of the CNNs are treated as hyper-parameters to be globally optimized. We cast the design of adaptive CNNs as a hyper-parameter optimization problem with respect to energy, accuracy, and communication constraints imposed by the mobile device. To efficiently solve this problem, we adapt Bayesian optimization to the properties of the design space, reaching near-optimal configurations in few tens of function evaluations. Our method reduces the energy consumed for image classification on a mobile device by up to 6×, compared to the best previously published work that uses CNNs as blackboxes. Finally, we evaluate two image classification practices, i.e., classifying all images locally versus over the cloud under energy and communication constraints.