使用Albu进行数据增强的时候，如果传入空gt_box_label会导致mmdet运行出现类型不匹配报错 got Long for the destination and Double for the source

[cnroge]

Thanks for your error report and we appreciate it a lot.

Checklist

1. I have searched related issues but cannot get the expected help.
2. I have read the FAQ documentation but cannot get the expected help.
3. The bug has not been fixed in the latest version.

Describe the bug
当我在mmdet中使用Albu的Pipeline类进行数据增强时候，出现了以下报错：
RuntimeError: Index put requires the source and destination dtypes match, got Long for the destination and Double for the source.
判定是在box_iou的时候生成的类别是int，而赋值的gt_box_label却是float

Error traceback

Traceback (most recent call last):
  File "tools/train.py", line 121, in <module>
    main()
  File "tools/train.py", line 117, in main
    runner.train()
  File "/home/cnroge/.conda/envs/ccccnroge/lib/python3.8/site-packages/mmengine/runner/runner.py", line 1777, in train
    model = self.train_loop.run()  # type: ignore
  File "/home/cnroge/.conda/envs/ccccnroge/lib/python3.8/site-packages/mmengine/runner/loops.py", line 287, in run
    self.run_iter(data_batch)
  File "/home/cnroge/.conda/envs/ccccnroge/lib/python3.8/site-packages/mmengine/runner/loops.py", line 311, in run_iter
    outputs = self.runner.model.train_step(
  File "/home/cnroge/.conda/envs/ccccnroge/lib/python3.8/site-packages/mmengine/model/wrappers/distributed.py", line 121, in train_step
    losses = self._run_forward(data, mode='loss')
  File "/home/cnroge/.conda/envs/ccccnroge/lib/python3.8/site-packages/mmengine/model/wrappers/distributed.py", line 161, in _run_forward
    results = self(**data, mode=mode)
  File "/home/cnroge/.conda/envs/ccccnroge/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "/home/cnroge/.conda/envs/ccccnroge/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl
    return forward_call(*args, **kwargs)
  File "/home/cnroge/.conda/envs/ccccnroge/lib/python3.8/site-packages/torch/nn/parallel/distributed.py", line 1519, in forward
    else self._run_ddp_forward(*inputs, **kwargs)
  File "/home/cnroge/.conda/envs/ccccnroge/lib/python3.8/site-packages/torch/nn/parallel/distributed.py", line 1355, in _run_ddp_forward
    return self.module(*inputs, **kwargs)  # type: ignore[index]
  File "/home/cnroge/.conda/envs/ccccnroge/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "/home/cnroge/.conda/envs/ccccnroge/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl
    return forward_call(*args, **kwargs)
  File "/home/cnroge/mmdetection/mmdet/models/detectors/base.py", line 92, in forward
    return self.loss(inputs, data_samples)
  File "/home/cnroge/mmdetection/mmdet/models/detectors/two_stage.py", line 174, in loss
    rpn_losses, rpn_results_list = self.rpn_head.loss_and_predict(
  File "/home/cnroge/mmdetection/mmdet/models/dense_heads/base_dense_head.py", line 165, in loss_and_predict
    losses = self.loss_by_feat(*loss_inputs)
  File "/home/cnroge/mmdetection/mmdet/models/dense_heads/rpn_head.py", line 125, in loss_by_feat
    losses = super().loss_by_feat(
  File "/home/cnroge/mmdetection/mmdet/models/dense_heads/anchor_head.py", line 502, in loss_by_feat
    cls_reg_targets = self.get_targets(
  File "/home/cnroge/mmdetection/mmdet/models/dense_heads/anchor_head.py", line 378, in get_targets
    results = multi_apply(
  File "/home/cnroge/mmdetection/mmdet/models/utils/misc.py", line 219, in multi_apply
    return tuple(map(list, zip(*map_results)))
  File "/home/cnroge/mmdetection/mmdet/models/dense_heads/anchor_head.py", line 252, in _get_targets_single
    assign_result = self.assigner.assign(pred_instances, gt_instances,
  File "/home/cnroge/mmdetection/mmdet/models/task_modules/assigners/max_iou_assigner.py", line 234, in assign
    assign_result = self.assign_wrt_overlaps(overlaps, gt_labels)
  File "/home/cnroge/mmdetection/mmdet/models/task_modules/assigners/max_iou_assigner.py", line 318, in assign_wrt_overlaps
    assigned_labels[pos_inds] = gt_labels[assigned_gt_inds[pos_inds] -
RuntimeError: Index put requires the source and destination dtypes match, got Long for the destination and Double for the source.

Bug fix
经过排查，发现是数据在经过Albu处理后，如果是空gt_bboxes_labels则会被转换成float64，以下是对应转换的代码处：

mmdet.datasets.transforms.transforms.py
class Albu(BaseTransform):
        ...
        def _postprocess_results(...):
                ...
(line1774)results[label] = np.array(
                        [results[label][i] for i in results['idx_mapper']])

我将1774行代码改为了

results[label] = np.array(
                        [results[label][i] for i in results['idx_mapper']], dtype=np.int64)

保证了numpy转换前后的数据类型一致即可