Real-time Inference with 2D Convolutional Neural Networks on Field Programmable Gate Arrays for High-rate Particle Imaging Detectors

TL;DR

This paper demonstrates a custom, optimized 2D CNN implementation on FPGA hardware for real-time particle detector data selection, focusing on accuracy, latency, and resource efficiency for future neutrino experiments.

Contribution

It introduces a resource-efficient, quantized 2D CNN optimized with KerasTuner and deployed on FPGA using hls4ml for real-time particle imaging data selection.

Findings

Achieved optimized accuracy and latency on FPGA

Validated feasibility for DUNE data acquisition system

Demonstrated effective network quantization for resource minimization

Abstract

We present a custom implementation of a 2D Convolutional Neural Network (CNN) as a viable application for real-time data selection in high-resolution and high-rate particle imaging detectors, making use of hardware acceleration in high-end Field Programmable Gate Arrays (FPGAs). To meet FPGA resource constraints, a two-layer CNN is optimized for accuracy and latency with KerasTuner, and network \textit{quantization} is further used to minimize the computing resource utilization of the network. We use "High Level Synthesis for Machine Learning" (\textit{hls4ml}) tools to test CNN deployment on a Xilinx UltraScale+ FPGA, which is a proposed FPGA technology for the front-end readout system of the future Deep Underground Neutrino Experiment (DUNE) far detector. We evaluate network accuracy and estimate latency and hardware resource usage, and comment on the feasibility of applying CNNs for real-time data selection within the proposed DUNE data acquisition system.