Hashing for Secure Optical Information Compression in a Heterogeneous Convolutional Neural Network
In the recent years, heterogeneous machine learning accelerators have become of significant interest in science, engineering and industry. The major processing speed bottlenecks in these platforms come from (a) an electronic data interconnect; (b) an electro-optical interface update rate. In this light, information compression implemented in native to incoming data optical domain could mitigate both problems mentioned above by reducing the demand on data throughput at the camera side and beyond. In this paper we present an optical hashing and compression scheme that is based on SWIFFT - a post-quantum hashing family of algorithms. High degree optical hardware-to-algorithm homomorphism allows to optimally harvest well-understood potential of free-space processing: innate parallelism, low latency tensor by-element multiplication and Fourier transform. The algorithm can provide several orders of magnitude increase in processing speed by replacing slow high-resolution CMOS cameras with ultra fast and signal-triggered CMOS detector arrays. Additionally, the information acquired in this way will require much lower transmission throughput, less \emph{in silico} processing power, storage, and will be pre-hashed facilitating cheap optical information security. This technology has a potential to allow heterogeneous convolutional 4f classifiers get closer in performance to their fully electronic counterparts.