A Camera That CNNs: Towards Embedded Neural Networks on Pixel Processor Arrays

TL;DR

This paper demonstrates a novel implementation of convolutional neural networks directly on pixel processor array sensors, enabling in-sensor image processing and recognition tasks without external hardware.

Contribution

It introduces a method to perform CNN operations directly on PPA hardware, including ternary weight convolutions and max-pooling, pioneering embedded neural network processing at the sensor level.

Findings

Successfully implemented CNN inference on PPA hardware

Achieved digit recognition and tracking tasks with embedded CNNs

Demonstrated real-time in-sensor image processing capabilities

Abstract

We present a convolutional neural network implementation for pixel processor array (PPA) sensors. PPA hardware consists of a fine-grained array of general-purpose processing elements, each capable of light capture, data storage, program execution, and communication with neighboring elements. This allows images to be stored and manipulated directly at the point of light capture, rather than having to transfer images to external processing hardware. Our CNN approach divides this array up into 4x4 blocks of processing elements, essentially trading-off image resolution for increased local memory capacity per 4x4 "pixel". We implement parallel operations for image addition, subtraction and bit-shifting images in this 4x4 block format. Using these components we formulate how to perform ternary weight convolutions upon these images, compactly store results of such convolutions, perform max-pooling, and transfer the resulting sub-sampled data to an attached micro-controller. We train ternary weight filter CNNs for digit recognition and a simple tracking task, and demonstrate inference of these networks upon the SCAMP5 PPA system. This work represents a first step towards embedding neural network processing capability directly onto the focal plane of a sensor.