Note

You are reading the documentation for MMSelfSup 0.x, which will soon be deprecated by the end of 2022. We recommend you upgrade to MMSelfSup 1.0.0rc versions to enjoy fruitful new features and better performance brought by OpenMMLab 2.0. Check out the changelog, code and documentation of MMSelfSup 1.0.0rc for more details.

Source code for mmselfsup.models.algorithms.simclr

# Copyright (c) OpenMMLab. All rights reserved.
import torch

from ..builder import ALGORITHMS, build_backbone, build_head, build_neck
from ..utils import GatherLayer
from .base import BaseModel


@ALGORITHMS.register_module()
class SimCLR(BaseModel):
    """SimCLR.

    Implementation of `A Simple Framework for Contrastive Learning
    of Visual Representations <https://arxiv.org/abs/2002.05709>`_.

    Args:
        backbone (dict): Config dict for module of backbone.
        neck (dict): Config dict for module of deep features to compact feature
            vectors. Defaults to None.
        head (dict): Config dict for module of loss functions.
            Defaults to None.
    """

    def __init__(self, backbone, neck=None, head=None, init_cfg=None):
        super(SimCLR, self).__init__(init_cfg)
        self.backbone = build_backbone(backbone)
        assert neck is not None
        self.neck = build_neck(neck)
        assert head is not None
        self.head = build_head(head)

    @staticmethod
    def _create_buffer(N):
        """Compute the mask and the index of positive samples.

        Args:
            N (int): batch size.
        """
        mask = 1 - torch.eye(N * 2, dtype=torch.uint8).cuda()
        pos_ind = (torch.arange(N * 2).cuda(),
                   2 * torch.arange(N, dtype=torch.long).unsqueeze(1).repeat(
                       1, 2).view(-1, 1).squeeze().cuda())
        neg_mask = torch.ones((N * 2, N * 2 - 1), dtype=torch.uint8).cuda()
        neg_mask[pos_ind] = 0
        return mask, pos_ind, neg_mask

    def extract_feat(self, img):
        """Function to extract features from 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, **kwargs):
        """Forward computation during training.

        Args:
            img (list[Tensor]): A list of input images with shape
                (N, C, H, W). Typically these should be mean centered
                and std scaled.

        Returns:
            dict[str, Tensor]: A dictionary of loss components.
        """
        assert isinstance(img, list)
        img = torch.stack(img, 1)
        img = img.reshape(
            (img.size(0) * 2, img.size(2), img.size(3), img.size(4)))
        x = self.extract_feat(img)  # 2n
        z = self.neck(x)[0]  # (2n)xd
        z = z / (torch.norm(z, p=2, dim=1, keepdim=True) + 1e-10)
        z = torch.cat(GatherLayer.apply(z), dim=0)  # (2N)xd
        assert z.size(0) % 2 == 0
        N = z.size(0) // 2
        s = torch.matmul(z, z.permute(1, 0))  # (2N)x(2N)
        mask, pos_ind, neg_mask = self._create_buffer(N)
        # remove diagonal, (2N)x(2N-1)
        s = torch.masked_select(s, mask == 1).reshape(s.size(0), -1)
        positive = s[pos_ind].unsqueeze(1)  # (2N)x1
        # select negative, (2N)x(2N-2)
        negative = torch.masked_select(s, neg_mask == 1).reshape(s.size(0), -1)
        losses = self.head(positive, negative)
        return losses