面向互联医疗中移动端设备的深度模型结构化贝叶斯压缩
计算机视觉与模式识别
2019-02-15 v1
摘要
深度模型,典型如深度神经网络,拥有数百万参数,能准确分析医疗数据,但耗时较长。然而,能耗经济性与计算效率对于开发和部署移动端设备至关重要,而后者是互联医疗的主流趋势。
引用
@article{arxiv.1902.05429,
title = {Structured Bayesian Compression for Deep models in mobile enabled devices for connected healthcare},
author = {Sijia Chen and Bin Song and Xiaojiang Du and Nadra Guizani},
journal= {arXiv preprint arXiv:1902.05429},
year = {2019}
}