Support skeleton (#155)

* pose * pose * pose * pose * 你的提交消息 * pose * pose * Delete train1.sh * pretreatment * configs * pose * reference * Update gaittr.py * naming * naming * Update transform.py * update for datasets * update README * update name and README * update * Update transform.py
2023-09-27 16:20:00 +08:00
parent 853bb1821d
commit 2c29afadf3
41 changed files with 4251 additions and 12 deletions
@@ -0,0 +1,107 @@
+'''
+Modifed fromhttps://github.com/BNU-IVC/FastPoseGait/blob/main/fastposegait/modeling/losses/supconloss.py
+'''
+
+import torch.nn as nn
+import torch
+from .base import BaseLoss, gather_and_scale_wrapper
+
+class SupConLoss_Re(BaseLoss):
+    def __init__(self, temperature=0.01):
+        super(SupConLoss_Re, self).__init__()
+        self.train_loss = SupConLoss(temperature=temperature)
+    @gather_and_scale_wrapper
+    def forward(self, features, labels=None, mask=None):
+        loss = self.train_loss(features,labels)
+        self.info.update({
+            'loss': loss.detach().clone()})
+        return loss, self.info
+
+
+class SupConLoss(nn.Module):
+    """Supervised Contrastive Learning: https://arxiv.org/pdf/2004.11362.pdf.
+    It also supports the unsupervised contrastive loss in SimCLR"""
+    def __init__(self, temperature=0.01, contrast_mode='all',
+                 base_temperature=0.07):
+        super(SupConLoss, self).__init__()
+        self.temperature = temperature
+        self.contrast_mode = contrast_mode
+        self.base_temperature = base_temperature
+
+    def forward(self, features, labels=None, mask=None):
+        """Compute loss for model. If both `labels` and `mask` are None,
+        it degenerates to SimCLR unsupervised loss:
+        https://arxiv.org/pdf/2002.05709.pdf
+        Args:
+            features: hidden vector of shape [bsz, n_views, ...].
+            labels: ground truth of shape [bsz].
+            mask: contrastive mask of shape [bsz, bsz], mask_{i,j}=1 if sample j
+                has the same class as sample i. Can be asymmetric.
+        Returns:
+            A loss scalar.
+        """
+        device = (torch.device('cuda')
+                  if features.is_cuda
+                  else torch.device('cpu'))
+
+        if len(features.shape) < 3:
+            raise ValueError('`features` needs to be [bsz, n_views, ...],'
+                             'at least 3 dimensions are required')
+        if len(features.shape) > 3:
+            features = features.view(features.shape[0], features.shape[1], -1)
+
+        batch_size = features.shape[0]
+        if labels is not None and mask is not None:
+            raise ValueError('Cannot define both `labels` and `mask`')
+        elif labels is None and mask is None:
+            mask = torch.eye(batch_size, dtype=torch.float32).to(device)
+        elif labels is not None:
+            labels = labels.contiguous().view(-1, 1)
+            if labels.shape[0] != batch_size:
+                raise ValueError('Num of labels does not match num of features')
+            mask = torch.eq(labels, labels.T).float().to(device)
+        else:
+            mask = mask.float().to(device)
+
+        contrast_count = features.shape[1]
+        contrast_feature = torch.cat(torch.unbind(features, dim=1), dim=0)
+        if self.contrast_mode == 'one':
+            anchor_feature = features[:, 0]
+            anchor_count = 1
+        elif self.contrast_mode == 'all':
+            anchor_feature = contrast_feature
+            anchor_count = contrast_count
+        else:
+            raise ValueError('Unknown mode: {}'.format(self.contrast_mode))
+
+        # compute logits
+        anchor_dot_contrast = torch.div(
+            torch.matmul(anchor_feature, contrast_feature.T),
+            self.temperature)
+        # for numerical stability
+        logits_max, _ = torch.max(anchor_dot_contrast, dim=1, keepdim=True)
+        logits = anchor_dot_contrast - logits_max.detach()
+
+        # tile mask
+        mask = mask.repeat(anchor_count, contrast_count)
+        # mask-out self-contrast cases
+        logits_mask = torch.scatter(
+            torch.ones_like(mask),
+            1,
+            torch.arange(batch_size * anchor_count).view(-1, 1).to(device),
+            0
+        )
+        mask = mask * logits_mask
+
+        # compute log_prob
+        exp_logits = torch.exp(logits) * logits_mask
+        log_prob = logits - torch.log(exp_logits.sum(1, keepdim=True))
+
+        # compute mean of log-likelihood over positive
+        mean_log_prob_pos = (mask * log_prob).sum(1) / mask.sum(1)
+
+        # loss
+        loss = - (self.temperature / self.base_temperature) * mean_log_prob_pos
+        loss = loss.view(anchor_count, batch_size).mean()
+
+        return loss
@@ -0,0 +1,19 @@
+'''
+Modifed fromhttps://github.com/BNU-IVC/FastPoseGait/blob/main/fastposegait/modeling/losses/supconloss_Lp.py
+'''
+
+from .base import BaseLoss, gather_and_scale_wrapper
+from pytorch_metric_learning import losses, distances
+
+class SupConLoss_Lp(BaseLoss):
+    def __init__(self, temperature=0.01):
+        super(SupConLoss_Lp, self).__init__()
+        self.distance = distances.LpDistance()
+        self.train_loss = losses.SupConLoss(temperature=temperature, distance=self.distance)
+    @gather_and_scale_wrapper
+    def forward(self, features, labels=None, mask=None):
+        loss = self.train_loss(features,labels)
+        self.info.update({
+            'loss': loss.detach().clone()})
+        return loss, self.info
+