Deep learning as a service paradigms are increasingly employed for image-based applications spanning surveillance, healthcare, biometrics, and e-commerce. Typically, trained convolutional neural networks (CNNs) are hosted on cloud infrastructure, and applied for inference on input images. There is interest in approaches to
enhance data privacy and security in such settings. Fully homomorphic encryption (FHE) can address this need as it caters to computations on encrypted data, but poses intensive computational burden. Prior works have proposed approaches to alleviate this burden for 32 x 32 images, but practical applications require at
least 10X higher resolution. Here, we present CareNets: Compact and Resource Efficient CNN for homomorphic inference on encrypted high-resolution images. Our approach is based on a novel compact packing scheme that packs CNN inputs, weights and activations densely into HE ciphertexts; and integrates them into the CNN computation flow. We implement CareNets using a GPU-accelerated FHE library for CNN inference on encrypted retinal images of size 96 x 96 and 256 x 256. Our results show that CareNets achieves over 32.78 x speedup, 45 x improvement in memory efficiency, and 5851 x reduction in transferred message size while maintaining accuracy within 3% of the non-encrypted CNN baselines.
License type:
PublisherCopyrights
Funding Info:
This project was supported by funding from the Deep Learning 2.0 program at the Institute for
Infocomm Research (I2R), A*STAR, Singapore; research grants from the US National Institutes of
Health (NIH grants R01EY19474, P30EY010572, and K12EY027720) and the US National Science
Foundation (NSF grants SCH-1622679 and SCH-1622542); unrestricted departmental funding from
the Oregon Health Sciences University, and a Career Development Award from Research to Prevent
Blindness (New York, NY).
Description:
Publisher - NeurIPS Workshop on Privacy in Machine Learning (PriML 2019)