如何运行预训练的像素目标模型

Question

我无法运行Suyog Dutt Jain等人在2017年发布的像素对象性的预训练模型。在

我参考了中使用预训练模型的部分自述文件.md但无法很好地理解以下过程。在

・设置：从这里下载并安装Deeplab-v1

・参考演示.py有关如何运行代码的分步说明。在

我试着安装并运行”演示.py“，但我有一些错误。 虽然我在网站上提到了解决方案，但仍然存在无法解决的错误。在

/usr/bin/python2.7 /mnt/C6BF-F678/pixelobjectness/pixelobjectness-master/demo.py
/usr/local/DL-Box/digits-4.0/caffe/distribute/bin/caffe.bin test --model=/mnt/C6BF-F678/pixelobjectness/pixelobjectness-master/test.prototxt --weights=/mnt/C6BF-F678/pixelobjectness/pixelobjectness-master/pixel_objectness.caffemodel --gpu=0 --iterations=4
I0606 14:34:09.451035  9830 caffe.cpp:261] Use GPU with device ID 0
I0606 14:34:09.451524  9830 caffe.cpp:265] GPU device name: Quadro P6000
[libprotobuf ERROR google/protobuf/text_format.cc:245] Error parsing text-format caffe.NetParameter: 424:16: Message type "caffe.LayerParameter" has no field named "interp_param".
F0606 14:34:09.811586  9830 upgrade_proto.cpp:79] Check failed: ReadProtoFromTextFile(param_file, param) Failed to parse NetParameter file: /mnt/C6BF-F678/pixelobjectness/pixelobjectness-master/test.prototxt
*** Check failure stack trace: ***
@     0x7f3829e1cdaa  (unknown)
@     0x7f3829e1cce4  (unknown)
@     0x7f3829e1c6e6  (unknown)
@     0x7f3829e1f687  (unknown)
@     0x7f382a58ef7e  caffe::ReadNetParamsFromTextFileOrDie()
@     0x7f382a5c8418  caffe::Net<>::Net()
@           0x407954  test()
@           0x40625c  main
@     0x7f3828979f45  (unknown)
@           0x406a2d  (unknown)
@              (nil)  (unknown)
Aborted (core dumped)

Process finished with exit code 0

我已经安装了GPU和caffe。在

如果你能仔细解释程序，我将不胜感激。在