ZYNQ_Target-Recognition

描述：实现了一个卷积神经网络加速器，成功搭载Yolov3tiny。配合摄像头采集+显示器回显环路，构建了一个高性能实时目标识别与检测系统。

 

 

实现方式：

1. Verilog实现卷积加速器的设计，

2. C语言实现Zynq PS端的开发，

3. Python实现神经网络的搭建与量化

开发工具套件：

1. Vivado2019.2，

2. Vitis2019.2，

3. Python，

4. Pytorch

验证平台：Xilinx Zynq xc7z100

文件目录：

1. src文件夹是所有代码文件

2. sim文件夹是testbench文件

3. vivado_prj文件夹是卷积加速器硬件Vivado工程

4. vitis_prj文件夹是基于vivado_prj硬件工程的软件Vitis工程

 

 

github：testbench001/ZYNQ_Target-识别 (github.com)

本栏目推荐文章
• IDEA项目名称后面出现中括号，模块Modules的名子和文件夹名称不同，可以右键修改名称也可以在File->Project Structure 修改Modules的Name(快捷键ctrl+Shift+Alt+s)
• slam相关-Sophus库cmake报错target "CUDA::cusolver" not found.
• 源码学习 出现彻底解决unable to find valid certification path to requested target
• Vscode插件Project Manager管理项目器无法显示新建标签的解决方法
• keil 工程编译报错browse infomation of one or more files is not available: ……Doing a project rebuild might fix this.
• Uninstall or delete MariaDB completely for re-installation
• Kafka消费端抛出异常Offset commit cannot be completed since the consumer is not part of an active group for auto partition assignment; it is likely that the consumer was kicked out of the group的解决方案
• svn/git代码管理规范，多Target工程适用
• 【Idea・Spring】Maven形のSpringBootのprojectを手動で構築するマニュアル
• cocoaPod 执行 pod install 时出现警告：The `XX [Release]` target overrides the `CLANG_ALLOW_NON_MODULAR_INCLUDES_IN_FRAMEWORK_MODULES` build setting defined in `xcconfig'.

ZYNQ_Target-Recognition Recognition complete Project Targetzynq_target-recognition recognition complete zynq_target-recognition specified directory project target recognition face_recognition recognition-vgg length-insensitive insensitive recognition decoding clustering-based hierarchical personalized recognition recognition relation networks local recognition residual learning resnet