International Association
for Cryptologic Research

<p> Machine learning applications gain more and more access to highly sensitive information while simultaneously requiring more and more computation resources. Hence, the need for outsourcing these computational expensive tasks while still ensuring security and confidentiality of the data is imminent. In their seminal work, Tramer and Boneh presented the Slalom protocol for privacy-preserving inference by splitting the computation into a data-independent preprocessing phase and a very efficient online phase. In this work, we present a new method to significantly speed up the preprocessing phase by introducing the Carnival protocol. Carnival leverages the pseudo-randomness of the Subset sum problem to also enable efficient outsourcing during the preprocessing phase. In addition to a security proof we also include an empirical study analyzing the landscape of the uniformity of the output of the Subset sum function for smaller parameters. Our findings show that Carnival is a great candidate for real-world implementations. </p>

The adoption of machine learning solutions is rapidly increasing across all parts of society. As the models grow larger, both training and inference of machine learning models is increasingly outsourced, e.g. to cloud service providers. This means that potentially sensitive data is processed on untrusted platforms, which bears inherent data security and privacy risks. In this work, we investigate how to protect distributed machine learning systems, focusing on deep convolutional neural networks. The most common and best-performing mixed MPC approaches are based on HE, secret sharing, and garbled circuits. They commonly suffer from large performance overheads, big accuracy losses, and communication overheads that grow linearly in the depth of the neural network. To improve on these problems, we present Dash, a fast and distributed private convolutional neural network inference scheme secure against malicious attackers. Building on arithmetic garbling gadgets [BMR16] and fancy-garbling [BCM+19], Dash is based purely on arithmetic garbled circuits. We introduce LabelTensors that allow us to leverage the massive parallelity of modern GPUs. Combined with state-of-the-art garbling optimizations, Dash outperforms previous garbling approaches up to a factor of about 100. Furthermore, we introduce an efficient scaling operation over the residues of the Chinese remainder theorem representation to arithmetic garbled circuits, which allows us to garble larger networks and achieve much higher accuracy than previous approaches. Finally, Dash requires only a single communication round per inference step, regardless of the depth of the neural network, and a very small constant online communication volume.