Data-free quantization via mixed-precision compensation without fine-tuning
Authors:
- Jun ChenJCJun Chen
- Shipeng BaiSBShipeng Bai
- Tianxin HuangTHTianxin Huang
- Mengmeng WangMWMengmeng Wang
- Guanzhong TianGTGuanzhong Tian
- Yong LiuYLYong Liu
- 26 June 2023
- journal article
- research article
- Published by Elsevier in Pattern Recognition
- Vol. 143
Abstract
No abstract availableKeywords
Funding Information
- Shanxi Provincial Key Research and Development Project (2021C01035)
This publication has 8 references indexed in Scilit:
- Ambiguity-aware robust teacher (ART): Enhanced self-knowledge distillation framework with pruned teacher networkPattern Recognition, 2023
- Sensitivity pruner: Filter-Level compression algorithm for deep neural networksPattern Recognition, 2023
- Bayesian asymmetric quantized neural networksPattern Recognition, 2023
- Computation-Efficient Knowledge Distillation via Uncertainty-Aware MixupPattern Recognition, 2023
- Mixed-precision quantized neural networks with progressively decreasing bitwidthPattern Recognition, 2021
- A Learning Framework for n-Bit Quantized Neural Networks Toward FPGAsIEEE Transactions on Neural Networks and Learning Systems, 2020
- ImageNet classification with deep convolutional neural networksCommunications of the ACM, 2017
- Data-free Parameter Pruning for Deep Neural Networks2015