https://doi.org/10.1051/epjconf/202328713020
Multi-channel free space optical convolutions
1 Max Planck Institute for Medical Research, Jahnstraße 29, 69120 Heidelberg
2 Institute for Molecular Systems Engineering and Advanced Materials, Universität Heidelberg, Neuenheimer Feld 225, 69120 Heidel-berg
† Equal Contribution
* e-mail: alexander.song@mr.mpg.de
** e-mail: nikhilesh.kottapalli@mr.mpg.de
*** e-mail: peer.fischer@mr.mpg.de
Published online: 18 October 2023
Convolutional layers are a critical feature of modern neural networks and require significant com-putational resources. In response, optical accelerators have been developed as a low-energy, high-bandwidth approach for performing large-scale convolutions. We extend these methods to act on many input channels each with their own set of convolutional kernels. We simulate the performance of this system with ray-tracing and evaluate its performance.
© The Authors, published by EDP Sciences, 2023
This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
