#batched-nms

Three detections, two dogs and one cat, are post-processed with torchvision:

import torch
from torchvision.ops import batched_nms

boxes = torch.tensor([[10., 10., 110., 110.],   # dog
                      [20., 20., 120., 120.],   # cat
                      [15., 12., 112., 115.]])  # dog
scores = torch.tensor([0.9, 0.8, 0.7])
labels = torch.tensor([0, 1, 0])               # 0 = dog, 1 = cat
keep = batched_nms(boxes, scores, labels, iou_threshold=0.5)
print(keep.tolist())

The pairwise IoUs are 0.68 (boxes 0 and 1), 0.87 (boxes 0 and 2) and 0.78 (boxes 1 and 2). What is printed?