image_segmenter

Image Segmentation using torchvision's pretrained fcn_resnet_101_coco model.

• Download the pre-trained fcn_resnet_101_coco image segmentation model's state_dict from the following URL :

https://download.pytorch.org/models/fcn_resnet101_coco-7ecb50ca.pth

wget https://download.pytorch.org/models/fcn_resnet101_coco-7ecb50ca.pth

• Create a model archive file and serve the fcn model in TorchServe using below commands

  torch-model-archiver --model-name fcn_resnet_101 --version 1.0 --model-file examples/image_segmenter/fcn/model.py --serialized-file fcn_resnet101_coco-7ecb50ca.pth --handler image_segmenter --extra-files examples/image_segmenter/fcn/fcn.py,examples/image_segmenter/fcn/intermediate_layer_getter.py
  mkdir model_store
  mv fcn_resnet_101.mar model_store/
  torchserve --start --model-store model_store --models fcn=fcn_resnet_101.mar
  curl http://127.0.0.1:8080/predictions/fcn -T examples/image_segmenter/fcn/persons.jpg

• Output

[[[11.49452    11.49452    11.49452    ... 10.846567   10.846567
   10.846567  ]
  [11.49452    11.49452    11.49452    ... 10.846567   10.846567
   10.846567  ]
  [11.49452    11.49452    11.49452    ... 10.846567   10.846567
   10.846567  ]
  ...
  [10.028987   10.028987   10.028987   ...  9.980104    9.980104
    9.980104  ]
  [10.028987   10.028987   10.028987   ...  9.980104    9.980105
    9.980105  ]
  [10.028987   10.028987   10.028987   ...  9.980104    9.980105
    9.980105  ]]

 [[-2.524181   -2.524181   -2.524181   ... -1.4157648  -1.4157648
   -1.4157648 ]
  [-2.524181   -2.524181   -2.524181   ... -1.4157648  -1.4157648
   -1.4157648 ]
  [-2.524181   -2.524181   -2.524181   ... -1.4157648  -1.4157648
   -1.4157648 ]
  ...
  [-0.52271044 -0.52271044 -0.52271044 ... -0.910931   -0.910931
   -0.910931  ]
  [-0.52271044 -0.52271044 -0.52271044 ... -0.910931   -0.910931
   -0.910931  ]
  [-0.52271044 -0.52271044 -0.52271044 ... -0.910931   -0.910931
   -0.910931  ]]

 [[-1.1300591  -1.1300591  -1.1300591  ... -0.88538504 -0.88538504
   -0.88538504]
  [-1.1300591  -1.1300591  -1.1300591  ... -0.88538504 -0.88538504
   -0.88538504]
  [-1.1300591  -1.1300591  -1.1300591  ... -0.88538504 -0.88538504
   -0.88538504]
  ...
  [-1.1726367  -1.1726367  -1.1726367  ... -1.6144376  -1.6144376
   -1.6144376 ]
  [-1.1726367  -1.1726367  -1.1726367  ... -1.6144376  -1.6144376
   -1.6144376 ]
  [-1.1726367  -1.1726367  -1.1726367  ... -1.6144376  -1.6144376
   -1.6144376 ]]

 ...

 [[-0.27236405 -0.27236405 -0.27236405 ... -0.791381   -0.79138106
   -0.79138106]
  [-0.27236405 -0.27236405 -0.27236405 ... -0.791381   -0.79138106
   -0.79138106]
  [-0.27236405 -0.27236405 -0.27236405 ... -0.791381   -0.79138106
   -0.79138106]
  ...
  [-0.16645516 -0.16645516 -0.16645516 ...  0.4377911   0.4377911
    0.4377911 ]
  [-0.16645516 -0.16645516 -0.16645516 ...  0.4377911   0.4377911
    0.4377911 ]
  [-0.16645516 -0.16645516 -0.16645516 ...  0.4377911   0.4377911
    0.4377911 ]]

 [[-0.70757735 -0.70757735 -0.70757735 ... -1.0088179  -1.0088179
   -1.0088179 ]
  [-0.70757735 -0.70757735 -0.70757735 ... -1.0088179  -1.0088179
   -1.0088179 ]
  [-0.70757735 -0.70757735 -0.70757735 ... -1.0088179  -1.0088179
   -1.0088179 ]
  ...
  [ 0.44139242  0.44139242  0.44139242 ...  0.28535858  0.28535858
    0.28535858]
  [ 0.44139242  0.44139242  0.44139242 ...  0.28535858  0.28535858
    0.28535858]
  [ 0.44139242  0.44139242  0.44139242 ...  0.28535858  0.28535858
    0.28535858]]

 [[-0.47415262 -0.47415262 -0.47415262 ... -0.4314881  -0.43148813
   -0.43148813]
  [-0.47415262 -0.47415262 -0.47415262 ... -0.4314881  -0.43148813
   -0.43148813]
  [-0.47415262 -0.47415262 -0.47415262 ... -0.4314881  -0.43148813
   -0.43148813]
  ...
  [ 0.04289126  0.04289126  0.04289126 ... -0.42638034 -0.42638034
   -0.42638034]
  [ 0.04289126  0.04289126  0.04289126 ... -0.42638034 -0.42638034
   -0.42638034]
  [ 0.04289126  0.04289126  0.04289126 ... -0.42638034 -0.42638034]