mmselfsup/openselfsup/models/rotation_pred.py

95 lines
3.2 KiB
Python

import torch
import torch.nn as nn
from openselfsup.utils import print_log
from . import builder
from .registry import MODELS
@MODELS.register_module
class RotationPred(nn.Module):
"""Rotation prediction.
Implementation of "Unsupervised Representation Learning
by Predicting Image Rotations (https://arxiv.org/abs/1803.07728)".
Args:
backbone (dict): Config dict for module of backbone ConvNet.
head (dict): Config dict for module of loss functions. Default: None.
pretrained (str, optional): Path to pre-trained weights. Default: None.
"""
def __init__(self, backbone, head=None, pretrained=None):
super(RotationPred, self).__init__()
self.backbone = builder.build_backbone(backbone)
if head is not None:
self.head = builder.build_head(head)
self.init_weights(pretrained=pretrained)
def init_weights(self, pretrained=None):
"""Initialize the weights of model.
Args:
pretrained (str, optional): Path to pre-trained weights.
Default: None.
"""
if pretrained is not None:
print_log('load model from: {}'.format(pretrained), logger='root')
self.backbone.init_weights(pretrained=pretrained)
self.head.init_weights(init_linear='kaiming')
def forward_backbone(self, img):
"""Forward backbone.
Args:
img (Tensor): Input images of shape (N, C, H, W).
Typically these should be mean centered and std scaled.
Returns:
tuple[Tensor]: backbone outputs.
"""
x = self.backbone(img)
return x
def forward_train(self, img, rot_label, **kwargs):
"""Forward computation during training.
Args:
img (Tensor): Input images of shape (N, C, H, W).
Typically these should be mean centered and std scaled.
rot_label (Tensor): Labels for the rotations.
kwargs: Any keyword arguments to be used to forward.
Returns:
dict[str, Tensor]: A dictionary of loss components.
"""
x = self.forward_backbone(img)
outs = self.head(x)
loss_inputs = (outs, rot_label)
losses = self.head.loss(*loss_inputs)
return losses
def forward_test(self, img, **kwargs):
x = self.forward_backbone(img) # tuple
outs = self.head(x)
keys = ['head{}'.format(i) for i in range(len(outs))]
out_tensors = [out.cpu() for out in outs] # NxC
return dict(zip(keys, out_tensors))
def forward(self, img, rot_label=None, mode='train', **kwargs):
if mode != "extract" and img.dim() == 5: # Nx4xCxHxW
assert rot_label.dim() == 2 # Nx4
img = img.view(
img.size(0) * img.size(1), img.size(2), img.size(3),
img.size(4)) # (4N)xCxHxW
rot_label = torch.flatten(rot_label) # (4N)
if mode == 'train':
return self.forward_train(img, rot_label, **kwargs)
elif mode == 'test':
return self.forward_test(img, **kwargs)
elif mode == 'extract':
return self.forward_backbone(img)
else:
raise Exception("No such mode: {}".format(mode))