Selective inference for accelerating deep learning-based image classification
Hyun-Yong Lee, Byung‐Tak Lee · 2016
Reducing the computational budget of inference in deep neural network while achieving high accuracy is important for time-sensitive applications. In this paper, unlike other approaches that try to compress a large neural network to a neural network with a smaller number of parameters, we try to complete the image classification as early as possible. Adding a middle output layer, we try to complete the image classification at the middle output layer when Top1 confidence exceeds a predefined confidence threshold. We prove the feasibility of proposed approach based on experiment using Inception-v3 in TensorFlow.