CrypTFlow2: Practical 2-Party Secure Inference

We present CrypTFlow2, a cryptographic framework for secure inference over Deep Neural Networks (DNNs) using secure 2-party computation.2 protocols are both correct -- i.e., their outputs are bitwise to the cleartext execution -- and efficient -- they outperform the-of-the-art protocols in both latency and scale. At the core of2, we have new 2PC protocols for secure comparison and division, carefully to balance round and communication complexity for secure tasks. Using CrypTFlow2, we present the first secure inference over-scale DNNs like ResNet50 and DenseNet121. These DNNs are at least an of magnitude larger than those considered in the prior work of 2-party inference. Even on the benchmarks considered by prior work, CrypTFlow2 an order of magnitude less communication and 20x-30x less time than state-of-the-art.