<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[Amazing_Fly的博客]]></title><description><![CDATA[]]></description><link>https://blog.youkuaiyun.com/Amazing_Fly</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; Amazing_Fly]]></copyright><item><title><![CDATA[YOLO 系列学习总结]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/164124930</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/164124930</guid><author>Amazing_Fly</author><pubDate>Thu, 27 Aug 2026 20:53:31 +0800</pubDate><description><![CDATA[维度参考先验预设锚框(aw, ah)（k-means 聚类）网格中心锚点（无宽高先验）位置回归（sigmoid 限网格）anchor - l（不受限）尺度回归（log 空间）(r+l, b+t)（直接回归距离）编码目标bbox2dist：四边距离解码decodedist2bbox（加减）优点尺度先验帮助小目标无先验超参、结构简单、易配合端到端输入图像▼检测头（anchor-free：锚点 + 距离回归）]]></description><category></category></item><item><title><![CDATA[torch.grid_sample的使用]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/152819541</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/152819541</guid><author>Amazing_Fly</author><pubDate>Thu, 09 Oct 2025 20:54:18 +0800</pubDate><description><![CDATA[本文主要针对grid_sample输出结果与像素坐标对应值有差异展开实验，最终发现该问题是由于align_corners参数导致，因此针对该参数进行了对比试验。其他参数如mode， padding参数带来的影响比较容易看出，这里不做赘述。]]></description><category></category></item><item><title><![CDATA[解决Ultralytics-YOLOv8-seg 由pt文件导出onnx后精度损失]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/144111791</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/144111791</guid><author>Amazing_Fly</author><pubDate>Thu, 28 Nov 2024 15:09:42 +0800</pubDate><description><![CDATA[前言：使用Ultralytics 8.1.34中yolov8n-seg进行训练，并到处onnx模型后，onnx模型精度对比pt文件有明显损失，具体表现为检测到的分割区域会出现缺少情况。pt文件推理正常分割图片。onnx模型推理分割图片。]]></description><category></category></item><item><title><![CDATA[量化参数的确定——ShrinkRange]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/135912670</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/135912670</guid><author>Amazing_Fly</author><pubDate>Mon, 29 Jan 2024 15:29:53 +0800</pubDate><description><![CDATA[量化知识学习]]></description><category></category></item><item><title><![CDATA[MATLAB十进制转其它进制的代码]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/129367674</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/129367674</guid><author>Amazing_Fly</author><pubDate>Mon, 06 Mar 2023 17:58:41 +0800</pubDate><description><![CDATA[进制转换]]></description><category></category></item><item><title><![CDATA[window中的CMD命令]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/128950028</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/128950028</guid><author>Amazing_Fly</author><pubDate>Thu, 09 Feb 2023 11:34:42 +0800</pubDate><description><![CDATA[windows cmd命令]]></description><category></category></item><item><title><![CDATA[Git一些进阶的使用方法]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/126561855</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/126561855</guid><author>Amazing_Fly</author><pubDate>Sat, 27 Aug 2022 20:52:32 +0800</pubDate><description><![CDATA[git的一些进阶使用]]></description><category></category></item><item><title><![CDATA[Pytorch使用和技巧]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/126120617</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/126120617</guid><author>Amazing_Fly</author><pubDate>Tue, 02 Aug 2022 14:15:56 +0800</pubDate><description><![CDATA[pytorch学习笔记]]></description><category></category></item><item><title><![CDATA[torch使用总结]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/126031713</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/126031713</guid><author>Amazing_Fly</author><pubDate>Thu, 28 Jul 2022 11:44:32 +0800</pubDate><description><![CDATA[torch学习笔记]]></description><category></category></item><item><title><![CDATA[git连接远程仓库]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/125627313</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/125627313</guid><author>Amazing_Fly</author><pubDate>Tue, 05 Jul 2022 20:47:43 +0800</pubDate><description><![CDATA[git使用]]></description><category></category></item><item><title><![CDATA[tmux的简单使用]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/125607913</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/125607913</guid><author>Amazing_Fly</author><pubDate>Mon, 04 Jul 2022 20:46:05 +0800</pubDate><description><![CDATA[tmux使用]]></description><category></category></item><item><title><![CDATA[conda软件库安装]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/125539101</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/125539101</guid><author>Amazing_Fly</author><pubDate>Fri, 01 Jul 2022 14:43:23 +0800</pubDate><description><![CDATA[实习期间的笔记]]></description><category></category></item><item><title><![CDATA[Tensorflow加载和处理数据集——TFRecord格式]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/125172733</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/125172733</guid><author>Amazing_Fly</author><pubDate>Wed, 08 Jun 2022 13:49:06 +0800</pubDate><description><![CDATA[Tensorlfow学习笔记]]></description><category></category></item><item><title><![CDATA[Tensorflow加载和处理数据集——tensorflow数据集项目]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/125171070</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/125171070</guid><author>Amazing_Fly</author><pubDate>Tue, 07 Jun 2022 19:20:25 +0800</pubDate><description><![CDATA[tensorflow_datasets来加载和处理数据集]]></description><category></category></item><item><title><![CDATA[Tensorflow加载和处理数据集-基础篇]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/125081332</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/125081332</guid><author>Amazing_Fly</author><pubDate>Wed, 01 Jun 2022 13:02:53 +0800</pubDate><description><![CDATA[tensorflow数据加载和处理]]></description><category></category></item><item><title><![CDATA[LaTex常用公式整理]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/123844970</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/123844970</guid><author>Amazing_Fly</author><pubDate>Wed, 30 Mar 2022 14:22:31 +0800</pubDate><description><![CDATA[在公式下方输入字符
$$\hat \beta = \underset{\beta}{\arg \min} L(\beta)$$
$$\hat \beta = \arg \min \limits_{\beta} L(\beta)$$

效果：

数学模式
在LaTeX数学模式中，公式有两种形式——行内公式和行间公式。前者公式嵌入在行内，适用于简单短小的公式；后者居中独占一行，适用于比较长或重要的公式。公式中的空格均会被忽略，可以使用命令\quad或\qquad实现
在行间公式中，命令\tag{n}可以进行手动]]></description><category></category></item><item><title><![CDATA[RBF（径向基）神经网络]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/122941604</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/122941604</guid><author>Amazing_Fly</author><pubDate>Tue, 15 Feb 2022 12:03:49 +0800</pubDate><description><![CDATA[转载：https://www.cnblogs.com/pinking/p/9349695.html
只要模型是一层一层的，并使用AD/BP算法，就能称作 BP神经网络。RBF 神经网络是其中一个特例。本文主要包括以下内容：

什么是径向基函数
RBF神经网络
RBF神经网络的学习问题
RBF神经网络与BP神经网络的区别
RBF神经网络与SVM的区别
为什么高斯核函数就是映射到高维区间
前馈网络、递归网络和反馈网络
完全内插法


...]]></description><category></category></item><item><title><![CDATA[MATLAB批量将figure文件转为png图片]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/122915530</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/122915530</guid><author>Amazing_Fly</author><pubDate>Sun, 13 Feb 2022 22:56:26 +0800</pubDate><description><![CDATA[使用函数包括
dir()函数浏览指定路径下所有的文件
path = "" % 绝对路径或者相对路径
filelist = dir(path) % filelist中以结构体形式存放path路径下的所有文件名，结构体结构如下
filelist = 

  包含以下字段的 31×1 struct 数组:

    name % 文件名
    folder % 文件的绝对路径
    date   % 写入时间
    bytes   % 字节数
    isdir    %是否为文件夹   是：1 否：0
 ]]></description><category></category></item><item><title><![CDATA[MATLAB中plot函数使用]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/122666827</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/122666827</guid><author>Amazing_Fly</author><pubDate>Mon, 24 Jan 2022 14:34:18 +0800</pubDate><description><![CDATA[随机生成一组数据：
x = [1:1:10];
y = rand(1, 10);

使用plot函数绘制图像
plot(x, y)


对图像属性进行设置，可以使用以下代码：
p = plot(x, y);
p.LineWidth = 2; %线宽为2
p.Color = 'r';   % 红色线条
p.Marker = '*' ; % 设置标记类型
p.MarkerSize = 6; %设置标记大小
p.MarkerEdgeColor = 'b'; %设置标记边缘颜色
p.MarkerFaceColor ]]></description><category></category></item><item><title><![CDATA[Tensorflow_gpu 2.0版安装记录]]></title><link>https://blog.youkuaiyun.com/Amazing_Fly/article/details/119459508</link><guid>https://blog.youkuaiyun.com/Amazing_Fly/article/details/119459508</guid><author>Amazing_Fly</author><pubDate>Fri, 06 Aug 2021 17:17:23 +0800</pubDate><description><![CDATA[忙了一下午，终于把Tensorflow GPU 2.4.0版本安装好了，Mark以下，以防以后出问题
Cuda, cudnn的安装可以参考b站的视频，讲解很详细。视频链接如下：https://www.bilibili.com/video/BV1HV411q7xD?p=6&amp;spm_id_from=pageDriver
只是其中有一些细节没有明确提出来，比如视频中说解压后将原来下载的cudnn的文件夹命名为“cudnn”放到CUDA问价夹下面，其实是将cudnn中的问价夹复制到CUDA下面，使其中文件补]]></description><category></category></item></channel></rss>