xuyunyunaixuexi的博客

1.下载安装cmake2.安装且永久配置opencv（debug+release）3.安装vs2017（开始电脑只有vs2013， 貌似不支持， 很多编译错误）mean, std很重要，开始直接原图没有预处理进行inference， 效果很差， 加了就ok了...std::shared_ptr<torch::jit::script::Module> module = ...

int main() { int img_size = 224; torch::DeviceType device_type = at::kCUDA; cout << "use cuda" << endl; // Deserialize the ScriptModule from a file using torch::jit::load(). Module ...

记代码中的一些小细节：解码代码：def decode_segmap(label_mask, dataset, plot=False): """Decode segmentation class labels into a color image Args: label_mask (np.ndarray): an (M,N) array of in...

from dataloaders import utilscriterion = utils.cross_entropy2d import torch.nn as nndef cross_entropy2d(logit, target, ignore_index=255, weight=None, size_average=True, batch_average=True): n...

记录下小tips:

最近遇到代码中的一个warining,UserWarning: nn.functional.upsample is deprecated. Use nn.functional.interpolate instead. warnings.warn("nn.functional.upsample is deprecated. Use nn.functional.interpolate inste...

https://github.com/jfzhang95/pytorch-deeplab-xceptionhttps://www.jianshu.com/p/026c5d78d3b1https://www.zhihu.com/question/282696986

class RandomScaleCrop(object): def __init__(self, base_size, crop_size, fill=0): self.base_size = base_size self.crop_size = crop_size self.fill = fill def __call__(self, ...

pytorch的自定义接口是真的方便， 记录一下自己分割数据输入的脚本：# -*- coding: utf-8 -*-# @Time : 2019/10/31 21:36# @Author : Yunyun Xu# @Contact : 1443563995@qq.com# @File : MyDatasetReader.py# @Software: Pycharm...

https://discuss.pytorch.org/t/seeking-help-the-model-does-not-appear-in-the-graphs-with-tensorboard/53572论坛中so many people都碰到这个问题了，大家可以看下论坛，有大佬给了解决方案但是我发现更简单的方法就是卸载你装的tensorboardpip uninstall te...

https://www.cnblogs.com/QingHuan/p/10016959.htmlnet = Net()if torch.cuda.is_available(): print("Using GPU")device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")net.to(devi...

pretrained_dict =...model_dict = model.state_dict()# 1. filter out unnecessary keyspretrained_dict = {k: v for k, vin pretrained_dict.items() if k inmodel_dict}# 2. overwrite entries in the ex...

import torch.utils.data as data#PIL: Python Image Library缩写，图像处理模块# Image,ImageFont,ImageDraw,ImageFilterfrom PIL import Image import osimport os.path # 图片扩展（图片格式）IMG_EXTENSIONS = [ ...

安装细节：1.cuda安装选择自定义安装比较好，且要将Driver components 应该勾上，英伟达公司为了方便，把cuda和相应的nvidia Driver打包在一起。否则就有CUDA driver version is insufficient for CUDA runtime version，是说driver驱动程序的版本太低，以致cuda不能运行。2.记录一下测试路径...