def __call__(self, class_logits):
"""
Computes the loss for Faster R-CNN.
This requires that the subsample method has been called beforehand.
Arguments:
class_logits (list[Tensor])
Returns:
classification_loss (Tensor)
"""
class_logits = cat(class_logits, dim=0)
device = class_logits.device
if not hasattr(self, "_proposal_pairs"):
raise RuntimeError("subsample needs to be called before")
proposals = self._proposal_pairs
labels = cat([proposal.get_field("labels") for proposal in proposals],
dim=0)
rel_fg_cnt = len(labels.nonzero())
rel_bg_cnt = labels.shape[0] - rel_fg_cnt
ce_weights = labels.new(class_logits.size(1)).fill_(1).float()
ce_weights[0] = float(rel_fg_cnt) / (rel_bg_cnt + 1e-5)
classification_loss = F.cross_entropy(class_logits,
labels,
weight=ce_weights)
return classification_loss
def __call__(self, class_logits):
"""
Computes the loss for Faster R-CNN.
This requires that the subsample method has been called beforehand.
Arguments:
class_logits (list[Tensor])
Returns:
classification_loss (Tensor)
"""
class_logits = cat(class_logits, dim=0)
device = class_logits.device
if not hasattr(self, "_proposal_pairs"):
raise RuntimeError("subsample needs to be called before")
proposals = self._proposal_pairs
labels = cat([proposal.get_field("labels") for proposal in proposals],
dim=0)
classification_loss = F.cross_entropy(class_logits, labels)
return classification_loss
def obj_classification_loss(self, proposals, class_logits):
class_logits = cat(class_logits, dim=0)
device = class_logits.device
labels = cat([proposal.get_field("labels") for proposal in proposals],
dim=0)
classification_loss = F.cross_entropy(class_logits, labels)
return classification_loss
def __call__(self, class_logits, box_regression):
"""
Computes the loss for Faster R-CNN.
This requires that the subsample method has been called beforehand.
Arguments:
class_logits (list[Tensor])
box_regression (list[Tensor])
Returns:
classification_loss (Tensor)
box_loss (Tensor)
"""
class_logits = cat(class_logits, dim=0)
box_regression = cat(box_regression, dim=0)
device = class_logits.device
if not hasattr(self, "_proposals"):
raise RuntimeError("subsample needs to be called before")
proposals = self._proposals
labels = cat([proposal.get_field("labels") for proposal in proposals],
dim=0)
regression_targets = cat([
proposal.get_field("regression_targets") for proposal in proposals
],
dim=0)
classification_loss = F.cross_entropy(class_logits, labels)
# get indices that correspond to the regression targets for
# the corresponding ground truth labels, to be used with
# advanced indexing
sampled_pos_inds_subset = torch.nonzero(labels > 0).squeeze(1)
labels_pos = labels[sampled_pos_inds_subset]
if self.cls_agnostic_bbox_reg:
map_inds = torch.tensor([4, 5, 6, 7], device=device)
else:
map_inds = 4 * labels_pos[:, None] + torch.tensor([0, 1, 2, 3],
device=device)
box_loss = smooth_l1_loss(
box_regression[sampled_pos_inds_subset[:, None], map_inds],
regression_targets[sampled_pos_inds_subset],
size_average=False,
beta=1,
)
box_loss = box_loss / labels.numel()
return classification_loss, box_loss
def pred_classification_loss(self, class_logits, freq_prior=None):
"""
Computes the loss for Faster R-CNN.
This requires that the subsample method has been called beforehand.
Arguments:
class_logits (list[Tensor])
Returns:
classification_loss (Tensor)
"""
class_logits = cat(class_logits, dim=0)
device = class_logits.device
if not hasattr(self, "_proposal_pairs"):
raise RuntimeError("subsample needs to be called before")
proposals = self._proposal_pairs
labels = cat([proposal.get_field("labels") for proposal in proposals],
dim=0)
rel_fg_cnt = len(labels.nonzero())
rel_bg_cnt = labels.shape[0] - rel_fg_cnt
ce_weights = labels.new(class_logits.size(1)).fill_(1).float()
ce_weights[0] = float(rel_fg_cnt) / (rel_bg_cnt + 1e-5)
classification_loss = F.cross_entropy(class_logits,
labels,
weight=ce_weights)
# add an auxilary loss to mine some positive relationship pairs
# class_probs = torch.log_softmax(class_logits[:, 1:], dim=-1)
# freq_probs = torch.softmax(freq_prior[:, 1:], dim=-1)
# klloss = F.kl_div(class_probs, freq_probs, reduction='batchmean')
#
# classification_loss += klloss
# class_probs = torch.softmax(class_logits, dim=-1).detach()
# freq_labels = freq_prior.argmax(1)
# pred_labels = class_probs[:, 1:].argmax(1) + 1
# match_idx = (freq_labels == pred_labels).nonzero().view(-1)
# keep_idx = (labels[match_idx] == 0).nonzero().view(-1)
# match_idx = match_idx[keep_idx]
# if match_idx.numel() > 0:
# labels_mined = freq_labels[match_idx]
# class_logits_mined = class_logits[match_idx]
# # weights = labels.new(class_logits.size(0)).fill_(1).float()
# weights = class_probs.max(1)[0][match_idx].detach()
# classification_loss += (weights * F.cross_entropy(class_logits_mined, labels_mined, weight=ce_weights, reduction='none')).mean()
return classification_loss