def assign_wrt_areas(self,
is_bbox_in_gt_eff,
is_bbox_in_gt_ignore,
gt_labels=None,
gt_priority=None):
num_bboxes, num_gts = is_bbox_in_gt_eff.size(
0), is_bbox_in_gt_eff.size(1)
if gt_priority is None:
gt_priority = torch.arange(num_gts).to(is_bbox_in_gt_eff.device)
# the bigger, the more preferable to be assigned
assigned_gt_inds = is_bbox_in_gt_eff.new_full((num_bboxes, ),
0,
dtype=torch.long)
# ignored indices
inds_of_ignore = torch.any(is_bbox_in_gt_ignore, dim=1)
assigned_gt_inds[inds_of_ignore] = -1
if is_bbox_in_gt_eff.sum() == 0: # No gt match
return AssignResult(num_gts, assigned_gt_inds, None, labels=None)
bbox_priority = is_bbox_in_gt_eff.new_full((num_bboxes, num_gts),
-1,
dtype=torch.long)
inds_of_match = torch.any(
is_bbox_in_gt_eff,
dim=1) # whether the bbox is matched (to any gt)
# Each bbox could match with multiple gts.
# The following codes deals with this
matched_bbox_and_gt_correspondence = is_bbox_in_gt_eff[
inds_of_match] # shape [nmatch, k]
matched_bbox_gt_inds = torch.nonzero(
matched_bbox_and_gt_correspondence)[:, 1]
# the matched gt index of each positive bbox. shape [nmatch]
bbox_priority[is_bbox_in_gt_eff] = gt_priority[matched_bbox_gt_inds]
_, argmax_priority = bbox_priority[inds_of_match].max(
dim=1) # the maximum shape [nmatch]
# effective indices
assigned_gt_inds[inds_of_match] = argmax_priority + 1
if gt_labels is not None:
assigned_labels = assigned_gt_inds.new_zeros((num_bboxes, ))
pos_inds = torch.nonzero(assigned_gt_inds > 0).squeeze()
if pos_inds.numel() > 0:
assigned_labels[pos_inds] = gt_labels[
assigned_gt_inds[pos_inds] - 1]
else:
assigned_labels = None
return AssignResult(num_gts,
assigned_gt_inds,
None,
labels=assigned_labels)
def random(cls, rng=None, **kwargs):
"""
Args:
rng (None | int | numpy.random.RandomState): seed or state
Kwargs:
num_preds: number of predicted boxes
num_gts: number of true boxes
p_ignore (float): probability of a predicted box assinged to an
ignored truth
p_assigned (float): probability of a predicted box not being
assigned
p_use_label (float | bool): with labels or not
Returns:
AssignResult :
Example:
>>> from mmdet.core.bbox.samplers.sampling_result import * # NOQA
>>> self = SamplingResult.random()
>>> print(self.__dict__)
"""
from mmdet.core.bbox.samplers.random_sampler import RandomSampler
from mmdet.core.bbox.assigners.assign_result import AssignResult
from mmdet.core.bbox import demodata
rng = demodata.ensure_rng(rng)
# make probabalistic?
num = 32
pos_fraction = 0.5
neg_pos_ub = -1
assign_result = AssignResult.random(rng=rng, **kwargs)
# Note we could just compute an assignment
bboxes = demodata.random_boxes(assign_result.num_preds, rng=rng)
gt_bboxes = demodata.random_boxes(assign_result.num_gts, rng=rng)
if rng.rand() > 0.2:
# sometimes algorithms squeeze their data, be robust to that
gt_bboxes = gt_bboxes.squeeze()
bboxes = bboxes.squeeze()
if assign_result.labels is None:
gt_labels = None
else:
gt_labels = None # todo
if gt_labels is None:
add_gt_as_proposals = False
else:
add_gt_as_proposals = True # make probabalistic?
sampler = RandomSampler(num,
pos_fraction,
neg_pos_ubo=neg_pos_ub,
add_gt_as_proposals=add_gt_as_proposals,
rng=rng)
self = sampler.sample(assign_result, bboxes, gt_bboxes, gt_labels)
return self
def assign_wrt_inside(self, overlaps, gt_labels=None):
"""Assign w.r.t. the overlaps of bboxes with gts.
Args:
overlaps (Tensor): Overlaps between k gt_bboxes and n bboxes,
shape(k, n).
gt_labels (Tensor, optional): Labels of k gt_bboxes, shape (k, ).
None for binary classification
Returns:
:obj:`AssignResult`: The assign result.
"""
if overlaps.numel() == 0:
raise ValueError('No gt or proposals')
num_gts, num_bboxes = overlaps.size(0), overlaps.size(1)
# 1. assign -1 by default
assigned_gt_inds = overlaps.new_full((num_bboxes, ),
-1,
dtype=torch.long)
## for each anchor, which gt best overlaps with it
# for each anchor, the max iou of all gts
max_overlaps, argmax_overlaps = overlaps.max(dim=0)
## for each gt, which anchor best overlaps with it
# for each gt, the max iou of all proposals
gt_max_overlaps, gt_argmax_overlaps = overlaps.max(dim=1)
# 2. assign negative: below
neg_inds = (max_overlaps >= 0) & (max_overlaps <= self.neg_iou_thr)
assigned_gt_inds[neg_inds] = 0
# 3. assign positive: above positive IoU threshold
# 4. assign ignore: an anchor contains more than one gt center
pos_inds = max_overlaps >= self.pos_iou_thr
assigned_gt_inds[pos_inds] = argmax_overlaps[pos_inds] + 1
if gt_labels is not None:
assigned_labels = assigned_gt_inds.new_zeros((num_bboxes, ))
pos_inds = torch.nonzero(assigned_gt_inds > 0).squeeze()
if pos_inds.numel() > 0:
assigned_labels[pos_inds] = gt_labels[
assigned_gt_inds[pos_inds] - 1]
ignore_inds = torch.nonzero(assigned_labels < 0).squeeze()
assigned_gt_inds[ignore_inds] = -1
else:
assigned_labels = None
return AssignResult(num_gts,
assigned_gt_inds,
max_overlaps,
labels=assigned_labels)
def test_random_assign_result():
"""Test random instantiation of assign result to catch corner cases."""
from mmdet.core.bbox.assigners.assign_result import AssignResult
AssignResult.random()
AssignResult.random(num_gts=0, num_preds=0)
AssignResult.random(num_gts=0, num_preds=3)
AssignResult.random(num_gts=3, num_preds=3)
AssignResult.random(num_gts=0, num_preds=3)
AssignResult.random(num_gts=7, num_preds=7)
AssignResult.random(num_gts=7, num_preds=64)
AssignResult.random(num_gts=24, num_preds=3)
def assign_wrt_overlaps(self, overlaps, gt_labels=None):
"""Assign w.r.t. the overlaps of bboxes with gts.
Args:
overlaps (Tensor): Overlaps between k gt_bboxes and n bboxes,
shape(k, n).
gt_labels (Tensor, optional): Labels of k gt_bboxes, shape (k, ).
Returns:
:obj:`AssignResult`: The assign result.
"""
if overlaps.numel() == 0:
raise ValueError('No gt or proposals')
num_gts, num_bboxes = overlaps.size(0), overlaps.size(1)
# 1. assign -1 by default
assigned_gt_inds = overlaps.new_full((num_bboxes, ),
-1,
dtype=torch.long)
# for each anchor, which gt best overlaps with it
# for each anchor, the max iou of all gts
max_overlaps, argmax_overlaps = overlaps.max(dim=0)
# for each gt, which anchor best overlaps with it
# for each gt, the max iou of all proposals
gt_max_overlaps, gt_argmax_overlaps = overlaps.max(dim=1)
# 2. assign negative: below
if isinstance(self.neg_iou_thr, float):
assigned_gt_inds[(max_overlaps >= 0)
& (max_overlaps < self.neg_iou_thr)] = 0
elif isinstance(self.neg_iou_thr, tuple):
assert len(self.neg_iou_thr) == 2
assigned_gt_inds[(max_overlaps >= self.neg_iou_thr[0])
& (max_overlaps < self.neg_iou_thr[1])] = 0
# 3. assign positive: above positive IoU threshold
pos_inds = max_overlaps >= self.pos_iou_thr
assigned_gt_inds[pos_inds] = argmax_overlaps[pos_inds] + 1
# 4. assign fg: for each gt, proposals with highest IoU
for i in range(num_gts):
if gt_max_overlaps[i] > self.min_pos_iou:
if self.gt_max_assign_all:
max_iou_inds = overlaps[i, :] == gt_max_overlaps[i]
assigned_gt_inds[max_iou_inds] = i + 1
else:
assigned_gt_inds[gt_argmax_overlaps[i]] = i + 1
if gt_labels is not None:
assigned_labels = assigned_gt_inds.new_zeros((num_bboxes, ))
pos_inds = torch.nonzero(assigned_gt_inds > 0).squeeze()
if pos_inds.numel() > 0:
assigned_labels[pos_inds] = gt_labels[
assigned_gt_inds[pos_inds] - 1]
else:
assigned_labels = None
# print('assigned indexes', overlaps.size(), torch.sum(assigned_gt_inds>0))
# print('labels', assigned_labels)
return AssignResult(num_gts,
assigned_gt_inds,
max_overlaps,
labels=assigned_labels)
def assign(self,
bboxes,
num_level_bboxes,
gt_bboxes,
gt_bboxes_ignore=None,
gt_labels=None):
"""Assign gt to bboxes.
The assignment is done in following steps
1. compute iou between all bbox (bbox of all pyramid levels) and gt
2. compute center distance between all bbox and gt
3. on each pyramid level, for each gt, select k bbox whose center
are closest to the gt center, so we total select k*l bbox as
candidates for each gt
4. get corresponding iou for the these candidates, and compute the
mean and std, set mean + std as the iou threshold
5. select these candidates whose iou are greater than or equal to
the threshold as postive
6. limit the positive sample's center in gt
Args:
bboxes (Tensor): Bounding boxes to be assigned, shape(n, 4).
num_level_bboxes (List): num of bboxes in each level
gt_bboxes (Tensor): Groundtruth boxes, shape (k, 4).
gt_bboxes_ignore (Tensor, optional): Ground truth bboxes that are
labelled as `ignored`, e.g., crowd boxes in COCO.
gt_labels (Tensor, optional): Label of gt_bboxes, shape (k, ).
Returns:
:obj:`AssignResult`: The assign result.
"""
INF = 100000000
bboxes = bboxes[:, :4]
num_gt, num_bboxes = gt_bboxes.size(0), bboxes.size(0)
# compute iou between all bbox and gt
overlaps = self.iou_calculator(bboxes, gt_bboxes)
# assign 0 by default
assigned_gt_inds = overlaps.new_full((num_bboxes, ),
0,
dtype=torch.long)
if num_gt == 0 or num_bboxes == 0:
# No ground truth or boxes, return empty assignment
max_overlaps = overlaps.new_zeros((num_bboxes, ))
if num_gt == 0:
# No truth, assign everything to background
assigned_gt_inds[:] = 0
if gt_labels is None:
assigned_labels = None
else:
assigned_labels = overlaps.new_full((num_bboxes, ),
-1,
dtype=torch.long)
return AssignResult(num_gt,
assigned_gt_inds,
max_overlaps,
labels=assigned_labels)
# compute center distance between all bbox and gt
gt_cx = (gt_bboxes[:, 0] + gt_bboxes[:, 2]) / 2.0
gt_cy = (gt_bboxes[:, 1] + gt_bboxes[:, 3]) / 2.0
gt_points = torch.stack((gt_cx, gt_cy), dim=1)
bboxes_cx = (bboxes[:, 0] + bboxes[:, 2]) / 2.0
bboxes_cy = (bboxes[:, 1] + bboxes[:, 3]) / 2.0
bboxes_points = torch.stack((bboxes_cx, bboxes_cy), dim=1)
distances = (bboxes_points[:, None, :] -
gt_points[None, :, :]).pow(2).sum(-1).sqrt()
# Selecting candidates based on the center distance
candidate_idxs = []
start_idx = 0
for level, bboxes_per_level in enumerate(num_level_bboxes):
# on each pyramid level, for each gt,
# select k bbox whose center are closest to the gt center
end_idx = start_idx + bboxes_per_level
distances_per_level = distances[start_idx:end_idx, :]
selectable_k = min(self.topk, bboxes_per_level)
_, topk_idxs_per_level = distances_per_level.topk(selectable_k,
dim=0,
largest=False)
candidate_idxs.append(topk_idxs_per_level + start_idx)
start_idx = end_idx
candidate_idxs = torch.cat(candidate_idxs, dim=0)
# get corresponding iou for the these candidates, and compute the
# mean and std, set mean + std as the iou threshold
candidate_overlaps = overlaps[candidate_idxs, torch.arange(num_gt)]
overlaps_mean_per_gt = candidate_overlaps.mean(0)
overlaps_std_per_gt = candidate_overlaps.std(0)
overlaps_thr_per_gt = overlaps_mean_per_gt + overlaps_std_per_gt
is_pos = candidate_overlaps >= overlaps_thr_per_gt[None, :]
# limit the positive sample's center in gt
for gt_idx in range(num_gt):
candidate_idxs[:, gt_idx] += gt_idx * num_bboxes
ep_bboxes_cx = bboxes_cx.view(1, -1).expand(
num_gt, num_bboxes).contiguous().view(-1)
ep_bboxes_cy = bboxes_cy.view(1, -1).expand(
num_gt, num_bboxes).contiguous().view(-1)
candidate_idxs = candidate_idxs.view(-1)
# calculate the left, top, right, bottom distance between positive
# bbox center and gt side
l_ = ep_bboxes_cx[candidate_idxs].view(-1, num_gt) - gt_bboxes[:, 0]
t_ = ep_bboxes_cy[candidate_idxs].view(-1, num_gt) - gt_bboxes[:, 1]
r_ = gt_bboxes[:, 2] - ep_bboxes_cx[candidate_idxs].view(-1, num_gt)
b_ = gt_bboxes[:, 3] - ep_bboxes_cy[candidate_idxs].view(-1, num_gt)
is_in_gts = torch.stack([l_, t_, r_, b_], dim=1).min(dim=1)[0] > 0.01
is_pos = is_pos & is_in_gts
# if an anchor box is assigned to multiple gts,
# the one with the highest IoU will be selected.
overlaps_inf = torch.full_like(overlaps,
-INF).t().contiguous().view(-1)
index = candidate_idxs.view(-1)[is_pos.view(-1)]
overlaps_inf[index] = overlaps.t().contiguous().view(-1)[index]
overlaps_inf = overlaps_inf.view(num_gt, -1).t()
max_overlaps, argmax_overlaps = overlaps_inf.max(dim=1)
assigned_gt_inds[
max_overlaps != -INF] = argmax_overlaps[max_overlaps != -INF] + 1
if gt_labels is not None:
assigned_labels = assigned_gt_inds.new_full((num_bboxes, ), -1)
pos_inds = torch.nonzero(assigned_gt_inds > 0,
as_tuple=False).squeeze()
if pos_inds.numel() > 0:
assigned_labels[pos_inds] = gt_labels[
assigned_gt_inds[pos_inds] - 1]
else:
assigned_labels = None
return AssignResult(num_gt,
assigned_gt_inds,
max_overlaps,
labels=assigned_labels)
