本文提出一种结合模型分区与损失特征编码的边缘主机分区方法，针对资源受限的物联网平台，通过压缩模型分区后的中间数据，实现高效的数据传输。
Edge-Host Partitioning of Deep Neural Networks with Feature Space Encoding
for Resource-Constrained Internet-of-Things Platforms
It propose an edge-host partitioning method, which combines model partition with lossy feature encoding. That is, the intermediate data after model partition will be compressed using lossy feature encoding before transmission
扫一扫
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
