You May Also Like... Privacy: Recommendation Systems Meet PIR

Vadapalli, Adithya; Bayatbabolghani, Fattaneh; Henry, Ryan

doi:10.2478/popets-2021-0059

Cited by 4 publications

(7 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A recent work by Vadapalli et al [37] also use spline to compute reciprocal square root and realize it using DPF (Distributed Point Function) [6]. Our work and [37] present a trade-off between online computation and key size. The online compute in LLAMA grows proportional to the number of intervals of the spline.…”

Section: Other Related Workmentioning

confidence: 93%

“…The online compute in LLAMA grows proportional to the number of intervals of the spline. On the other hand, the online compute of [37] grows exponentially with the input bitwidth. (For reciprocal square root with 16-bit inputs, in the online phase, LLAMA makes 1448 AES evaluations, and [37] makes 131072 AES evaluations.)…”

Section: Other Related Workmentioning

confidence: 99%

“…On the other hand, the online compute of [37] grows exponentially with the input bitwidth. (For reciprocal square root with 16-bit inputs, in the online phase, LLAMA makes 1448 AES evaluations, and [37] makes 131072 AES evaluations.) However, the key size in LLAMA is higher than that in [37].…”

Section: Other Related Workmentioning

confidence: 99%

“…(For reciprocal square root with 16-bit inputs, in the online phase, LLAMA makes 1448 AES evaluations, and [37] makes 131072 AES evaluations.) However, the key size in LLAMA is higher than that in [37]. (For reciprocal square root with 16-bit inputs, key size in LLAMA is nearly 5KB, compared to around 0.3KB in [37].)…”

Section: Other Related Workmentioning

confidence: 99%

“…However, the key size in LLAMA is higher than that in [37]. (For reciprocal square root with 16-bit inputs, key size in LLAMA is nearly 5KB, compared to around 0.3KB in [37].) Since the primary benefit of FSS based techniques is to reduce online complexity, LLAMA would perform better than [37].…”

Section: Other Related Workmentioning

confidence: 99%

See 4 more Smart Citations

LLAMA: A Low Latency Math Library for Secure Inference

Gupta

Kumaraswamy

Chandran

et al. 2022

PoPETs

View full text Add to dashboard Cite

Secure machine learning (ML) inference can provide meaningful privacy guarantees to both the client (holding sensitive input) and the server (holding sensitive weights of the ML model) while realizing inferenceas-a-service. Although many specialized protocols exist for this task, including those in the preprocessing model (where a majority of the overheads are moved to an input independent offline phase), they all still suffer from large online complexity. Specifically, the protocol phase that executes once the parties know their inputs, has high communication, round complexity, and latency. Function Secret Sharing (FSS) based techniques offer an attractive solution to this in the trusted dealer model (where a dealer provides input independent correlated randomness to both parties), and 2PC protocols obtained based on these techniques have a very lightweight online phase. Unfortunately, current FSS-based 2PC works (AriaNN, PoPETS 2022; Boyle et al. Eurocrypt 2021; Boyle et al. TCC 2019) fall short of providing a complete solution to secure inference. First, they lack support for math functions (e.g., sigmoid, and reciprocal square root) and hence, are insufficient for a large class of inference algorithms (e.g. recurrent neural networks). Second, they restrict all values in the computation to be of the same bitwidth and this prevents them from benefitting from efficient float-to-fixed converters such as Tensorflow Lite that crucially use low bitwidth representations and mixed bitwidth arithmetic. In this work, we present LLAMA – an end-to-end, FSS based, secure inference library supporting precise low bitwidth computations (required by converters) as well as provably precise math functions; thus, overcoming all the drawbacks listed above. We perform an extensive evaluation of LLAMA and show that when compared with non-FSS based libraries supporting mixed bitwidth arithmetic and math functions (SIRNN, IEEE S&P 2021), it has at least an order of magnitude lower communication, rounds, and runtimes. We integrate LLAMA with the EzPC framework (IEEE EuroS&P 2019) and demonstrate its robustness by evaluating it on large benchmarks (such as ResNet-50 on the ImageNet dataset) as well as on benchmarks considered in AriaNN – here too LLAMA outperforms prior work.

show abstract

Section: Other Related Workmentioning

confidence: 93%

Section: Other Related Workmentioning

confidence: 99%