Many applications of mobile deep learning, especially real-time computer vision workloads, are constrained by computation power. This is particularly true for workloads running on older consumer phones, where a typical device might be powered by a single- or dual-core ARMv7 CPU. We provide an open-source implementation and a comprehensive analysis of (to our knowledge) the state of the art ultra-low-precision (<4 bit precision) implementation of the core primitives required for modern deep learning workloads on ARMv7 devices, and demonstrate speedups of 4x-20x over our additional state-of-the-art float32 and int8 baselines.
@article{arxiv.1712.02427,
title = {High performance ultra-low-precision convolutions on mobile devices},
author = {Andrew Tulloch and Yangqing Jia},
journal= {arXiv preprint arXiv:1712.02427},
year = {2017}
}
Comments
Presented at NIPS 2017, Machine Learning on the Phone and other Consumer Devices workshop