Hardware-aware One-Shot Neural Architecture Search in Coordinate Ascent Framework

Designing accurate and efficient convolutional neural architectures for vast amount of hardware is challenging because hardware designs are complex and diverse. This paper addresses the hardware diversity challenge in Neural Architecture Search (NAS). Unlike previous approaches that apply search algorithms on a small, human-designed search space without considering hardware diversity, we propose HURRICANE that explores the automatic hardware-aware search over a much larger search space and a multistep search scheme in coordinate ascent framework, to generate tailored models for different types of hardware. Extensive experiments on ImageNet show that our algorithm consistently achieves a much lower inference latency with a similar or better accuracy than state-of-the-art NAS methods on three types of hardware. Remarkably, HURRICANE achieves a 76.63% top-1 accuracy on ImageNet with a inference latency of only 16.5 ms for DSP, which is a 3.4% higher accuracy and a 6.35x inference speedup than FBNet-iPhoneX. For VPU, HURRICANE achieves a 0.53% higher top-1 accuracy than Proxyless-mobile with a 1.49x speedup. Even for well-studied mobile CPU, HURRICANE achieves a 1.63% higher top-1 accuracy than FBNet-iPhoneX with a comparable inference latency. HURRICANE also reduces the training time by 54.7% on average compared to SinglePath-Oneshot.