Deep learning inference time guarantee for aggregated images at the edge of the network

Image inference using deep learning (DL) has recently become popular in a variety of applications. In the near future, real-time for delay sensitive images will be widespread with edge computing paradigm, where DL processing executed at network. this study, method future is proposed to infer large number images, which are aggregated from various end devices toward an server and temporarily stored queue, different requirements on time. With method, time including waiting queue image guaranteed that tolerant minimized by adaptively partitioning between cloud depending arrival conditions server. The effectiveness evaluated via simulation.