Flash: A Hybrid Private Inference Protocol for Deep CNNs with High Accuracy and Low Latency on CPU. (arXiv:2401.16732v1 [cs.CR])

This paper presents Flash, an optimized private inference (PI) hybrid
protocol utilizing both homomorphic encryption (HE) and secure two-party
computation (2PC), which can reduce the end-to-end PI latency for deep CNN
models less than 1 minute with CPU. To this end, first, Flash proposes a
low-latency convolution algorithm built upon a fast slot rotation operation and
a novel data encoding scheme, which results in 4-94x performance gain over the
state-of-the-art. Second, to minimize the communication cost introduced by the
standard …

accuracy arxiv can cnn cnns computation cpu encryption end end-to-end flash high homomorphic encryption hybrid latency low party private protocol