arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.16024cs.AR
收藏

面向边缘AI硬件的少样本学习位宽感知设计环境

Bit-Width-Aware Design Environment for Few-Shot Learning on Edge AI Hardware

Research Institute of Electrical Communication, Tohoku University(东北大学电气通信研究所);Graduate School of Engineering, Tohoku University(东北大学工学研究科);IMT Atlantique, Lab-STICC, UMR CNRS 6285(大西洋高等理工学院,Lab-STICC,CNRS联合研究实验室6285)

R. Kanda, H. L. Blevec, N. Onizawa, M. Leonardon, V. Gripon, T. Hanyu

AI总结:

本文提出一种支持任意位宽的少样本学习设计环境,通过优化和调整提升边缘AI硬件的吞吐量和精度

摘要:

在本研究中,我们提出了一种在小型FPGA SoC(如PYNQ-Z1板)上实现实时少样本学习的实现方法,支持任意固定点位宽。传统的基于Tensil的设计环境将硬件实现限制在16或32位的固定点位宽。为解决这一问题,我们采用了FINN框架,使实现能够支持任意位宽。进行了若干定制和微调,包括:1.优化Transpose节点以解决数据格式不匹配问题,2.增加将最终的reduce mean操作转换为全局平均池化(GAP)的处理。这些调整使我们能够在保持与传统实现相同精度的情况下减少位宽,并在使用CIFAR-10数据集评估时实现约两倍的吞吐量。

英文摘要:

In this study, we propose an implementation methodology of real-time few-shot learning on tiny FPGA SoCs such as the PYNQ-Z1 board with arbitrary fixed-point bit-widths. Tensil-based conventional design environments limited hardware implementations to fixed-point bit-widths of 16 or 32 bits. To address this, we adopt the FINN framework, enabling implementations with arbitrary bit-widths. Several customizations and minor adjustments are made, including: 1.Optimization of Transpose nodes to resolve data format mismatches, 2.Addition of handling for converting the final reduce mean operation to Global Average Pooling (GAP). These adjustments allow us to reduce the bit-width while maintaining the same accuracy as the conventional realization, and achieve approximately twice the throughput in evaluations using CIFAR-10 dataset.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑