def test_iou_0_dim_cuda(self):
boxes1 = torch.rand(0, 5, dtype=torch.float32)
boxes2 = torch.rand(10, 5, dtype=torch.float32)
expected_ious = torch.zeros(0, 10, dtype=torch.float32)
ious_cuda = pairwise_iou_rotated(boxes1.cuda(), boxes2.cuda())
self.assertTrue(torch.allclose(ious_cuda.cpu(), expected_ious))
boxes1 = torch.rand(10, 5, dtype=torch.float32)
boxes2 = torch.rand(0, 5, dtype=torch.float32)
expected_ious = torch.zeros(10, 0, dtype=torch.float32)
ious_cuda = pairwise_iou_rotated(boxes1.cuda(), boxes2.cuda())
self.assertTrue(torch.allclose(ious_cuda.cpu(), expected_ious))
def test_iou_precision(self):
for device in ["cpu"] + ["cuda"] if torch.cuda.is_available() else []:
boxes1 = torch.tensor([[565, 565, 10, 10.0, 0]], dtype=torch.float32, device=device)
boxes2 = torch.tensor([[565, 565, 10, 8.3, 0]], dtype=torch.float32, device=device)
iou = 8.3 / 10.0
expected_ious = torch.tensor([[iou]], dtype=torch.float32)
ious = pairwise_iou_rotated(boxes1, boxes2)
self.assertTrue(torch.allclose(ious.cpu(), expected_ious))
def pairwise_iou(boxes1: RotatedBoxes, boxes2: RotatedBoxes) -> None:
"""
Given two lists of rotated boxes of size N and M,
compute the IoU (intersection over union)
between __all__ N x M pairs of boxes.
The box order must be (x_center, y_center, width, height, angle).
Args:
boxes1, boxes2 (RotatedBoxes):
two `RotatedBoxes`. Contains N & M rotated boxes, respectively.
Returns:
Tensor: IoU, sized [N,M].
"""
return pairwise_iou_rotated(boxes1.tensor, boxes2.tensor)
def softnms_rotated(boxes, scores, sigma, score_threshold, soft_mode="gaussian"):
assert soft_mode in ["linear", "gaussian"]
iou_matrix = pairwise_iou_rotated(boxes, boxes)
undone_mask = scores >= score_threshold
while undone_mask.sum() > 1:
idx = scores[undone_mask].argmax()
idx = undone_mask.nonzero(as_tuple=False)[idx].item()
undone_mask[idx] = False
ious = iou_matrix[idx, undone_mask]
scales = scale_by_iou(ious, sigma, soft_mode)
scores[undone_mask] *= scales
undone_mask[scores < score_threshold] = False
return scores
def test_iou_extreme(self):
# Cause floating point issues in cuda kernels (#1266)
# See: https://github.com/facebookresearch/detectron2/issues/1266
for device in ["cpu"] + ["cuda"] if torch.cuda.is_available() else []:
boxes1 = torch.tensor([[160.0, 153.0, 230.0, 23.0, -37.0]], device=device)
boxes2 = torch.tensor(
[
[
-1.117407639806935e17,
1.3858420478349148e18,
1000.0000610351562,
1000.0000610351562,
1612.0,
]
],
device=device,
)
ious = pairwise_iou_rotated(boxes1, boxes2)
self.assertTrue(ious.min() >= 0, ious)
def test_iou_too_many_boxes_cuda(self):
s1, s2 = 5, 1289035
boxes1 = torch.zeros(s1, 5)
boxes2 = torch.zeros(s2, 5)
ious_cuda = pairwise_iou_rotated(boxes1.cuda(), boxes2.cuda())
self.assertTupleEqual(tuple(ious_cuda.shape), (s1, s2))
def test_iou_half_overlap_cuda(self):
boxes1 = torch.tensor([[0.5, 0.5, 1.0, 1.0, 0.0]], dtype=torch.float32)
boxes2 = torch.tensor([[0.25, 0.5, 0.5, 1.0, 0.0]], dtype=torch.float32)
expected_ious = torch.tensor([[0.5]], dtype=torch.float32)
ious_cuda = pairwise_iou_rotated(boxes1.cuda(), boxes2.cuda())
self.assertTrue(torch.allclose(ious_cuda.cpu(), expected_ious))
def bench():
for _ in range(n):
pairwise_iou_rotated(b1, b2)
if dev.type == "cuda":
torch.cuda.synchronize()