tripletloss.py

import torch
from torch import nn


def normalize(x, axis=-1):
    """Normalizing to unit length along the specified dimension.
    Args:
      x: pytorch Variable
    Returns:
      x: pytorch Variable, same shape as input
    """
    x = 1. * x / (torch.norm(x, 2, axis, keepdim=True).expand_as(x) + 1e-12)
    return x


def euclidean_dist(x, y):
    """
    Args:
      x: pytorch Variable, with shape [m, d]
      y: pytorch Variable, with shape [n, d]
    Returns:
      dist: pytorch Variable, with shape [m, n]
    """
    m, n = x.size(0), y.size(0)
    xx = torch.pow(x, 2).sum(1, keepdim=True).expand(m, n)
    yy = torch.pow(y, 2).sum(1, keepdim=True).expand(n, m).t()
    dist = xx + yy
    dist.addmm_(1, -2, x, y.t())
    dist = dist.clamp(min=1e-12).sqrt()  # for numerical stability
    return dist


def hard_example_mining(dist_mat, labels, mask=None, return_inds=False):
    """For each anchor, find the hardest positive and negative sample.
    Args:
      dist_mat: pytorch Variable, pair wise distance between samples, shape [N, N]
      labels: pytorch LongTensor, with shape [N]
      mask: pytorch Tensor, with shape [N, N]
      return_inds: whether to return the indices. Save time if `False`(?)
    Returns:
      dist_ap: pytorch Variable, distance(anchor, positive); shape [N]
      dist_an: pytorch Variable, distance(anchor, negative); shape [N]
      p_inds: pytorch LongTensor, with shape [N];
        indices of selected hard positive samples; 0 <= p_inds[i] <= N - 1
      n_inds: pytorch LongTensor, with shape [N];
        indices of selected hard negative samples; 0 <= n_inds[i] <= N - 1
    NOTE: Only consider the case in which all labels have same num of samples,
      thus we can cope with all anchors in parallel.
    """

    assert len(dist_mat.size()) == 2
    assert dist_mat.size(0) == dist_mat.size(1)
    N = dist_mat.size(0)

    # shape [N, N]
    is_pos = labels.expand(N, N).eq(labels.expand(N, N).t()).float()
    is_neg = labels.expand(N, N).ne(labels.expand(N, N).t()).float()

    # `dist_ap` means distance(anchor, positive)
    # both `dist_ap` and `relative_p_inds` with shape [N, 1]
    if mask is None:
        mask = torch.ones_like(dist_mat)

    aux_mat = torch.zeros_like(dist_mat)
    aux_mat[mask == 0] -= 10
    dist_mat = dist_mat + aux_mat

    dist_ap, relative_p_inds = torch.max(
        (dist_mat * is_pos).contiguous().view(N, -1), 1, keepdim=True)

    # `dist_an` means distance(anchor, negative)
    # both `dist_an` and `relative_n_inds` with shape [N, 1]
    # dist_mat[dist_mat == 0] += 10000  # 处理非法值。归一化后的最大距离为2
    aux_mat = torch.zeros_like(dist_mat)
    aux_mat[mask == 0] += 10000
    dist_mat = dist_mat + aux_mat
    dist_an, relative_n_inds = torch.min(
        (dist_mat * is_neg).contiguous().view(N, -1), 1, keepdim=True)
    # shape [N]

    dist_ap = dist_ap.squeeze(1)
    dist_an = dist_an.squeeze(1)

    if return_inds:
        # shape [N, N]
        ind = (labels.new().resize_as_(labels)
               .copy_(torch.arange(0, N).long())
               .unsqueeze(0).expand(N, N))
        # shape [N, 1]
        p_inds = torch.gather(
            ind[is_pos].contiguous().view(N, -1), 1, relative_p_inds.data)
        n_inds = torch.gather(
            ind[is_neg].contiguous().view(N, -1), 1, relative_n_inds.data)
        # shape [N]
        p_inds = p_inds.squeeze(1)
        n_inds = n_inds.squeeze(1)
        return dist_ap, dist_an, p_inds, n_inds

    return dist_ap, dist_an


class TripletLoss(object):
    """Modified from Tong Xiao's open-reid (https://github.com/Cysu/open-reid).
    Related Triplet Loss theory can be found in paper 'In Defense of the Triplet
    Loss for Person Re-Identification'."""

    def __init__(self, margin=None):
        self.margin = margin
        if margin is not None:
            self.ranking_loss = nn.MarginRankingLoss(margin=margin)
        else:
            self.ranking_loss = nn.SoftMarginLoss()

    def __call__(self, global_feat, labels, mask=None, normalize_feature=False):
        """

        :param global_feat:
        :param labels:
        :param mask:  [N, N] 可见性mask。不可见的mask将不会被选择。若全部不可见，则对结果*0
        :param normalize_feature:
        :return:
        """
        if normalize_feature:
            global_feat = normalize(global_feat, axis=-1)
        dist_mat = euclidean_dist(global_feat, global_feat)
        dist_ap, dist_an = hard_example_mining(
            dist_mat, labels, mask=mask)
        y = dist_an.new().resize_as_(dist_an).fill_(1)
        if self.margin is not None:
            loss = self.ranking_loss(dist_an, dist_ap, y)
        else:
            loss = self.ranking_loss(dist_an - dist_ap, y)
        return loss  # , dist_ap, dist_an