def _generate_proposals(self, box):
""" Generates proposals by adding noise to the input box
args:
box - input box
returns:
torch.Tensor - Array of shape (num_proposals, 4) containing proposals
torch.Tensor - Array of shape (num_proposals,) containing IoU overlap of each proposal with the input box. The
IoU is mapped to [-1, 1]
"""
# Generate proposals
num_proposals = self.proposal_params['boxes_per_frame']
proposal_method = self.proposal_params.get('proposal_method',
'default')
if proposal_method == 'default':
proposals = torch.zeros((num_proposals, 4))
gt_iou = torch.zeros(num_proposals)
for i in range(num_proposals):
proposals[i, :], gt_iou[i] = prutils.perturb_box(
box,
min_iou=self.proposal_params['min_iou'],
sigma_factor=self.proposal_params['sigma_factor'])
elif proposal_method == 'gmm':
proposals, _, _ = prutils.sample_box_gmm(
box,
self.proposal_params['proposal_sigma'],
num_samples=num_proposals)
gt_iou = prutils.iou(box.view(1, 4), proposals.view(-1, 4))
# Map to [-1, 1]
gt_iou = gt_iou * 2 - 1
return proposals, gt_iou
def _generate_proposals(self, box, rng):
""" Generates proposals by adding noise to the input box
args:
box - input box
returns:
array - Array of shape (num_proposals, 4) containing proposals
array - Array of shape (num_proposals,) containing IoU overlap of each proposal with the input box. The
IoU is mapped to [-1, 1]
"""
# Generate proposals
num_proposals = self.proposal_params['boxes_per_frame']
proposals = np.zeros((num_proposals, 4))
gt_iou = np.zeros(num_proposals)
for i in range(num_proposals):
proposals[i, :], gt_iou[i] = prutils.perturb_box(
box,
min_iou=self.proposal_params['min_iou'],
sigma_factor=self.proposal_params['sigma_factor'],
rng=rng)
# Map to [-1, 1]
gt_iou = gt_iou * 2 - 1
return proposals, gt_iou
