Mz的博客

-- coding:utf-8 --import osimport caffe as caffeimport cv2 as cvimport numpy as npfrom scipy import interpolatefrom scipy.optimize import brentqfrom sklearn import metricsfrom tqdm import tqdmimport shutilimport csvfrom evaluate import lfwlfw_p

“”“Functions for building the face recognition network.“””MIT LicenseCopyright © 2016 David SandbergPermission is hereby granted, free of charge, to any person obtaining a copyof this software and associated documentation files (the “Software”), to

# _*_ coding:UTF-8 _*_#开发作者 : ZhangRong z00520111#开发时间 : 2020/5/13 10:09#文件名称 : firetest.py#开发工具 : PyCharm#Description:#Copyright @ Huawei Technologies Co., Ltd. 2019-2020. All rights reserved.import cv2import osimport numpy as np

需要修改的地方是红线部分，根据分类数量进行修改，原版的开源是21个分类（包括background），所以是63（213）,126（213*2）,如果class_num为5，则为15，30

1、Check failed: datum_channels > 0 (0 vs. 0).train.prototxt中Datatype AnnotatedData和Data互换，changing data type from “AnnotatedData” to“Data” or “Data” to "AnnotatedData"layer { name: "Data1"...

http://megaface.cs.washington.edu/dataset/download.html

/etc/profile添加路径sudo vi /etc/profile，在最后面添加下面三句export PATH=$PATH:/usr/local/cuda-10.0/binexport LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda-10.0/lib64export LIBRARY_PATH=$LIBRARY_PATH:/usr/loc...

# -*- coding: utf-8 -*-import reimport requestsfrom pyquery import PyQuery as pq# from getcookie import excuteScriptimport time, randomimport jsonimport osrequests.packages.urllib3.disable_wa...

https://github.com/shicai/MobileNet-Caffe

# _*_ coding:UTF-8 _*_# 开发作者 : ZhangRong z00520111# 开发时间 : 2020/3/28 10:09# 文件名称 : catchhuoche.py# 开发工具 : PyCharm# Description:# Copyright @ Huawei Technologies Co., Ltd. 2019-2...

前言此篇博文，博主将详细介绍Jetson TX1如何安装运行caffe框架。因为要在开发板上跑ssd的demo，所以就直接clone ssd项目了，反正ssd也是caffe的一个分支，除了ssd特有的层，二者在编译运行上几乎一样。安装依赖包依次安装以下依赖项，如果不想每一次install都出现yes or no的提示项，可以在命令末尾加上“-y”。sudo apt-get install ...

## Refer to http://caffe.berkeleyvision.org/installation.html# Contributions simplifying and improving our build system are welcome!# cuDNN acceleration switch (uncomment to build with cuDNN). USE...

# 'superclean' target recursively* deletes all files ending with an extension# in $(SUPERCLEAN_EXTS) below. This may be useful if you've built older# versions of Caffe that do not place all genera...

##############################Derive compiler warning dump locations##############################WARNS_EXT := warnings.txtCXX_WARNS := $(addprefix $(BUILD_DIR)/, KaTeX parse error: Expected '}', ...

！！！

因为最近涉及到保存模型推理结果的输出文件，所以学一学如何将数据保存为CSV文件，比如保存检测框box的顶点，置信度，label，入侵检测的结果等。用到的也是C++的ofstream，ofstream有一个操作是"<<"，这个也很好用，就类似std::cout的操作一样即可。比如我现在的数据是比较统一的，每一个样本是一行，一行数据要分成四列，第一列是样本的图像地址，第二列是标签，第三列...

# -*- coding=utf-8 -*-# top 1 accuracy 0.9249791286257038 top k accuracy 0.9747623788455786import osimport randomimport tensorflow.contrib.slim as slimimport timeimport loggingimport numpy as n...