Abstract
Going deeper and wider in neural architectures improves their accuracy, while the limited GPU DRAM places an undesired restriction on the network desi......
小提示：本篇文献需要登录阅读全文，点击跳转登录
Full Text Link
