NeuMan是苹果公司开发的机器学习模型,可使用神经辐射场从单个视频中重建背景场景和动画人物。
环境使用Conda创建环境:condaenvcreate-fenvironment.yml或者,您可以通过执行以下命令来创建环境:condacreate-nneuman_envpython=3.7-y;condaactivateneuman_env;condainstallpytorch==1.8.0torchvision==0.9.0cudatoolkit=10.2-cpytorch;#ForRTX30seriesGPUwithCUDAversion11.x,pleaseuse:#condainstallpytorch==1.8.0torchvision==0.9.0torchaudio==0.8.0cudatoolkit=11.1-cpytorch-cconda-forgecondainstall-cfvcore-ciopath-cconda-forgefvcoreiopath;condainstall-cbottlernvidiacub;condainstallpytorch3d-cpytorch3d;condainstall-cconda-forgeigl;pipinstallopencv-pythonjoblibopen3dimageiotensorboardXchumpylpipsscikit-imageipythonmatplotlib;请注意,pytorch3d在例子中需要特定版本的pytorch pytorch=1.8.0。激活环境:condaactivateneuman_env点击空白处退出提示
评论