Survey: Exploiting Data Redundancy for Optimization of Deep Learning

Data redundancy is ubiquitous in the inputs and intermediate results of Deep Neural Networks (DNN) . It offers many significant opportunities for improving DNN performance efficiency has been explored a large body work. These studies have scattered venues across several years. The targets they focus on range from images to videos texts, techniques use detect exploit data also vary aspects. There not yet systematic examination summary efforts, making it difficult researchers get comprehensive view prior work, state art, differences shared principles, areas directions explore. This article tries fill void. surveys hundreds recent papers topic, introduces novel taxonomy put various into single categorization framework, description main methods used exploiting multiple kinds DNNs data, points out set research future exploration.