Fixed issue unpacking size in furthest_point_sample (#248)

The tuple unpacked into three arguments when four were given when
running VoteNet on sunrgbd

mmdet3d/ops/furthest_point_sample/furthest_point_sample.py

@@ -25,7 +25,7 @@ class FurthestPointSampling(Function):
         """
         assert points_xyz.is_contiguous()
 
-        B, N, _ = points_xyz.size()
+        B, N = points_xyz.size()[:2]
         output = torch.cuda.IntTensor(B, num_points)
         temp = torch.cuda.FloatTensor(B, N).fill_(1e10)