Cain: Automatic Code Generation for Simultaneous Convolutional Kernels on Focal-plane Sensor-processors
This work addresses the difficulty of programming FPSPs for edge computation, offering a tool to improve efficiency in low-power, high-frame-rate applications.
The authors tackled the challenge of developing complex algorithms for Focal-plane Sensor-processors (FPSPs) by creating Cain, a compiler that generates code for multiple convolutional kernels, resulting in code that is half as long compared to existing compilers for the SCAMP-5 FPSP.
Focal-plane Sensor-processors (FPSPs) are a camera technology that enable low power, high frame rate computation, making them suitable for edge computation. Unfortunately, these devices' limited instruction sets and registers make developing complex algorithms difficult. In this work, we present Cain - a compiler that targets SCAMP-5, a general-purpose FPSP - which generates code from multiple convolutional kernels. As an example, given the convolutional kernels for an MNIST digit recognition neural network, Cain produces code that is half as long, when compared to the other available compilers for SCAMP-5.