EBSP
Fangxin Liu, Wenbo Zhao, Zongwu Wang, Yongbiao Chen, Zhezhi He, Naifeng Jing, Xiaoyao Liang, Li Jiang
Abstract
Model compression has been extensively investigated for supporting efficient neural network inference on edge-computing platforms due to the huge model size and computation amount. Recent researches embrace joint-way compression across multiple techniques for extreme compression. However, most joint-way methods adopt a naive solution that applies two approaches sequentially, which can be sub-optimal, as it lacks a systematic approach to incorporate them.
Topics & Concepts
Computer scienceJoint (building)InferenceComputationData compressionCompression (physics)Enhanced Data Rates for GSM EvolutionArtificial neural networkArtificial intelligenceTheoretical computer scienceAlgorithmEngineeringComposite materialMaterials scienceArchitectural engineeringAdvanced Neural Network ApplicationsGaussian Processes and Bayesian InferenceMachine Learning and Algorithms