Picture for Jung Ho Ahn

Jung Ho Ahn

NeuJeans: Private Neural Network Inference with Joint Optimization of Convolution and Bootstrapping

Dec 07, 2023
Figure 1 for NeuJeans: Private Neural Network Inference with Joint Optimization of Convolution and Bootstrapping
Figure 2 for NeuJeans: Private Neural Network Inference with Joint Optimization of Convolution and Bootstrapping
Figure 3 for NeuJeans: Private Neural Network Inference with Joint Optimization of Convolution and Bootstrapping
Figure 4 for NeuJeans: Private Neural Network Inference with Joint Optimization of Convolution and Bootstrapping
Viaarxiv icon

HyPHEN: A Hybrid Packing Method and Optimizations for Homomorphic Encryption-Based Neural Networks

Feb 05, 2023
Figure 1 for HyPHEN: A Hybrid Packing Method and Optimizations for Homomorphic Encryption-Based Neural Networks
Figure 2 for HyPHEN: A Hybrid Packing Method and Optimizations for Homomorphic Encryption-Based Neural Networks
Figure 3 for HyPHEN: A Hybrid Packing Method and Optimizations for Homomorphic Encryption-Based Neural Networks
Figure 4 for HyPHEN: A Hybrid Packing Method and Optimizations for Homomorphic Encryption-Based Neural Networks
Viaarxiv icon

AESPA: Accuracy Preserving Low-degree Polynomial Activation for Fast Private Inference

Jan 18, 2022
Figure 1 for AESPA: Accuracy Preserving Low-degree Polynomial Activation for Fast Private Inference
Figure 2 for AESPA: Accuracy Preserving Low-degree Polynomial Activation for Fast Private Inference
Figure 3 for AESPA: Accuracy Preserving Low-degree Polynomial Activation for Fast Private Inference
Figure 4 for AESPA: Accuracy Preserving Low-degree Polynomial Activation for Fast Private Inference
Viaarxiv icon

Restructuring Batch Normalization to Accelerate CNN Training

Add code
Jul 04, 2018
Figure 1 for Restructuring Batch Normalization to Accelerate CNN Training
Figure 2 for Restructuring Batch Normalization to Accelerate CNN Training
Figure 3 for Restructuring Batch Normalization to Accelerate CNN Training
Figure 4 for Restructuring Batch Normalization to Accelerate CNN Training
Viaarxiv icon