a586351的博客

已知情况：你现在拥有一个以yolo格式标注的数据集yolo标签格式：“类别 x_center/img_width y_center/img_height width/img_width height/img_height”voc标签格式：“filename、size、xmin、ymin、xmax、ymax”转换代码：from xml.dom.minidom import Documentimport osimport os.pathfrom PIL import Imagean

环境：pyhton == 3.6dlib == 1.19.10scipyscikit-imagenumpyimutilsopencv环境安装：create - n face python=3.6conda activate faceconda install -c conda-forge dlib numpy=1.18.5 numpy-basepip install opencv-python3.4.5.20 opencv-contrib-python3.4.5.20 sci

环境:python3.6.8torch == 1.3.1torchvision == 0.4.1import torchimport torch.optim as optimimport torch.nn.functional as Fimport torchvisionimport torchvision.datasets as datasetsimport torchvis...

环境：Python3.6.8torch==1.3.1torchvision==0.4.1完整代码:import torchimport torch.optim as optimimport torch.nn.functional as Fimport torchvisionimport torchvision.datasets as datasetsimport torch...

首先说明我锁使用的是来自github的版本：https://github.com/eriklindernoren/PyTorch-YOLOv3大概也许会持续记录吧。。。我懒记于2019.11.07我的环境是Package VersionPillow 6.2.1pip 1...