Bugfix in BalancedPositiveNegativeSampler introduced during torchscript support (#1670)

test/test_models_detection_utils.py

+import torch
+from torchvision.models.detection import _utils
+import unittest
+
+
+class Tester(unittest.TestCase):
+    def test_balanced_positive_negative_sampler(self):
+        sampler = _utils.BalancedPositiveNegativeSampler(4, 0.25)
+        # keep all 6 negatives first, then add 3 positives, last two are ignore
+        matched_idxs = [torch.tensor([0, 0, 0, 0, 0, 0, 1, 1, 1, -1, -1])]
+        pos, neg = sampler(matched_idxs)
+        # we know the number of elements that should be sampled for the positive (1)
+        # and the negative (3), and their location. Let's make sure that they are
+        # there
+        self.assertEqual(pos[0].sum(), 1)
+        self.assertEqual(pos[0][6:9].sum(), 1)
+        self.assertEqual(neg[0].sum(), 3)
+        self.assertEqual(neg[0][0:6].sum(), 3)
+
+
+if __name__ == '__main__':
+    unittest.main()

torchvision/models/detection/_utils.py

@@ -11,10 +11,8 @@ import torchvision
 # TODO: https://github.com/pytorch/pytorch/issues/26727
 def zeros_like(tensor, dtype):
     # type: (Tensor, int) -> Tensor
-    if tensor.dtype == dtype:
-        return tensor.detach().clone()
-    else:
-        return tensor.to(dtype)
+    return torch.zeros_like(tensor, dtype=dtype, layout=tensor.layout,
+                            device=tensor.device, pin_memory=tensor.is_pinned())
 
 
 @torch.jit.script