/hls_cnn_accelerator

A Convolutional Neural Network Accelerator, which increases the process of convolution calculation. Based on Xilinx HLS design suite.

Primary LanguageC++

hls_cnn_accelerator

高级硬件设计(HLS)卷积神经网络加速器

一、完成功能

1.实现对多通道(16通道)卷积和池化计算进行加速;

2.实现卷积与池化的深度融合。

  1. 使用HLS工具,对其进行优化(如使用pipeline、partition等)。

二、实验记录

1.主要的实现在fpga_lab_2_cnn/conv.cpp中.测试代码为fpga_lab2_cnn/conv_test.cpp.