Integrated Photonic Fractional Convolution Accelerator

Abstract

An integrated photonic circuit architecture to perform a modified-convolution operation based on the Discrete Fractional Fourier Transform (DFrFT) is introduced. This is accomplished by utilizing two nonuniformly-coupled waveguide lattices with equally-spaced eigenmode spectra and with different lengths that perform DFrDT operations of complementary orders sandwiching a modulator array. Numerical simulations show that smoothing and edge detection tasks are indeed performed even for noisy input signals.

References in corpus (2)

• Efficient, Compact and Low Loss Thermo-Optic Phase Shifter in Silicon
• An Optical Frontend for a Convolutional Neural Network

Cited by in corpus (4)

• Learning Arbitrary Complex Matrices by Interlacing Amplitude and Phase Masks with Fixed Unitary Operations
• Integrated Photonic Programmable Random Matrix Generator with Minimal Active Components
• Photonic Matrix Multiplier Makes a Direction-Finding Sensor
• Embedding Matrices in Programmable Photonic Networks with Flexible Depth and Width