yj-fb/ckks_cryptonet

This project is a Python library that combines CKKS homomorphic encryption with CNN neural networks to enable secure, privacy-preserving inference for deep learning models. It extends the CryptoNet framework by adding enhanced debugging options, model export functionality, and improved API documentation. The library is designed for developers and researchers working on privacy-preserving machine learning who need to perform encrypted inference on sensitive data without exposing the underlying information.