Skip to content

longcw/RoIAlign.pytorch

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

37 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

RoIAlign for PyTorch

This is a PyTorch version of RoIAlign. This implementation is based on crop_and_resize and supports both forward and backward on CPU and GPU.

NOTE: Thanks meikuam for updating this repo for PyTorch 1.0. You can find the original version for torch <= 0.4.1 in pytorch_0.4 branch.

Introduction

The crop_and_resize function is ported from tensorflow, and has the same interface with tensorflow version, except the input feature map should be in NCHW order in PyTorch. They also have the same output value (error < 1e-5) for both forward and backward as we expected, see the comparision in test.py.

Note: Document of crop_and_resize can be found here. And RoIAlign is a wrap of crop_and_resize that uses boxes with unnormalized (x1, y1, x2, y2) as input (while crop_and_resize use normalized (y1, x1, y2, x2) as input). See more details about the difference of RoIAlign and crop_and_resize in tensorpack.

Warning: Currently it only works using the default GPU (index 0)

Usage

  • Install and test

    python setup.py install
    ./test.sh
    
  • Use RoIAlign or crop_and_resize

    Since PyTorch 1.2.0 Legacy autograd function with non-static forward method is deprecated. We use new-style autograd function with static forward method. Example:

    import torch
    from roi_align import RoIAlign      # RoIAlign module
    from roi_align import CropAndResize # crop_and_resize module
    
    # input feature maps (suppose that we have batch_size==2)
    image = torch.arange(0., 49).view(1, 1, 7, 7).repeat(2, 1, 1, 1)
    image[0] += 10
    print('image: ', image)
    
    
    # for example, we have two bboxes with coords xyxy (first with batch_id=0, second with batch_id=1).
    boxes = torch.Tensor([[1, 0, 5, 4],
                         [0.5, 3.5, 4, 7]])
    
    box_index = torch.tensor([0, 1], dtype=torch.int) # index of bbox in batch
    
    # RoIAlign layer with crop sizes:
    crop_height = 4
    crop_width = 4
    roi_align = RoIAlign(crop_height, crop_width)
    
    # make crops:
    crops = roi_align(image, boxes, box_index)
    
    print('crops:', crops)

    Output:

    image:  tensor([[[[10., 11., 12., 13., 14., 15., 16.],
          [17., 18., 19., 20., 21., 22., 23.],
          [24., 25., 26., 27., 28., 29., 30.],
          [31., 32., 33., 34., 35., 36., 37.],
          [38., 39., 40., 41., 42., 43., 44.],
          [45., 46., 47., 48., 49., 50., 51.],
          [52., 53., 54., 55., 56., 57., 58.]]],
    
    
        [[[ 0.,  1.,  2.,  3.,  4.,  5.,  6.],
          [ 7.,  8.,  9., 10., 11., 12., 13.],
          [14., 15., 16., 17., 18., 19., 20.],
          [21., 22., 23., 24., 25., 26., 27.],
          [28., 29., 30., 31., 32., 33., 34.],
          [35., 36., 37., 38., 39., 40., 41.],
          [42., 43., 44., 45., 46., 47., 48.]]]])
          
    crops: tensor([[[[11.0000, 12.0000, 13.0000, 14.0000],
              [18.0000, 19.0000, 20.0000, 21.0000],
              [25.0000, 26.0000, 27.0000, 28.0000],
              [32.0000, 33.0000, 34.0000, 35.0000]]],
    
    
            [[[24.5000, 25.3750, 26.2500, 27.1250],
              [30.6250, 31.5000, 32.3750, 33.2500],
              [36.7500, 37.6250, 38.5000, 39.3750],
              [ 0.0000,  0.0000,  0.0000,  0.0000]]]])