我目前正在开发一个具有自定义关键点数据集的项目。 在进一步处理数据集之前，我希望检查数据集和训练过程是否正确。
我遇到了这个错误，尝试在 Google 和 StackOverflow 中查找类似的问题，我发现很难通过浏览来定位问题。
我只使用总数据集中的 5 个来检查性能。 它有 18 个以 COCO 格式注释的关键点。
from detectron2.data.datasets import register_coco_instances
register_coco_instances("train_t", {}, "/content/drive/MyDrive/thesis/test/point_test-2.json", "/content/drive/MyDrive/thesis/test/train")
register_coco_instances("val_t", {}, "/content/drive/MyDrive/thesis/test/point_test-2.json", "/content/drive/MyDrive/thesis/test/val")
sample_metadata = MetadataCatalog.get("train_t")
dataset_dicts = DatasetCatalog.get("train_t")
from detectron2.data import MetadataCatalog
keypoint_names = ['0', '1', '2', '3', '4', '5','6','7','8','9',
'10', '11', '12', '13', '14', '15','16','17']
keypoint_flip_map = [('0', '1'), ('2', '15'), ('3','4'), ('5','6'),('7','8'),('9','10'),
('11', '12'), ('13', '14'), ('16','17')]
classes = MetadataCatalog.get("train_t").thing_classes = ['points']
print(classes)
MetadataCatalog.get("train_t").thing_classes = ['points']
MetadataCatalog.get("train_t").thing_dataset_id_to_contiguous_id = {1:0}
MetadataCatalog.get("train_t").keypoint_names = keypoint_names
MetadataCatalog.get("train_t").keypoint_flip_map = keypoint_flip_map
MetadataCatalog.get("train_t").evaluator_type="coco"
这是我遇到的错误，
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
<ipython-input-13-cb06f7a198f0> in <module>()
29 trainer = DefaultTrainer(cfg) #CocoTrainer(cfg)
30 trainer.resume_or_load(resume=False)
---> 31 trainer.train()
8 frames
/usr/local/lib/python3.7/dist-packages/detectron2/engine/defaults.py in train(self)
482 OrderedDict of results, if evaluation is enabled. Otherwise None.
483 """
--> 484 super().train(self.start_iter, self.max_iter)
485 if len(self.cfg.TEST.EXPECTED_RESULTS) and comm.is_main_process():
486 assert hasattr(
/usr/local/lib/python3.7/dist-packages/detectron2/engine/train_loop.py in train(self, start_iter, max_iter)
147 for self.iter in range(start_iter, max_iter):
148 self.before_step()
--> 149 self.run_step()
150 self.after_step()
151 # self.iter == max_iter can be used by `after_train` to
/usr/local/lib/python3.7/dist-packages/detectron2/engine/defaults.py in run_step(self)
492 def run_step(self):
493 self._trainer.iter = self.iter
--> 494 self._trainer.run_step()
495
496 def state_dict(self):
/usr/local/lib/python3.7/dist-packages/detectron2/engine/train_loop.py in run_step(self)
265 If you want to do something with the data, you can wrap the dataloader.
266 """
--> 267 data = next(self._data_loader_iter)
268 data_time = time.perf_counter() - start
269
/usr/local/lib/python3.7/dist-packages/detectron2/data/common.py in __iter__(self)
232
233 def __iter__(self):
--> 234 for d in self.dataset:
235 w, h = d["width"], d["height"]
236 bucket_id = 0 if w > h else 1
/usr/local/lib/python3.7/dist-packages/torch/utils/data/dataloader.py in __next__(self)
519 if self._sampler_iter is None:
520 self._reset()
--> 521 data = self._next_data()
522 self._num_yielded += 1
523 if self._dataset_kind == _DatasetKind.Iterable and \
/usr/local/lib/python3.7/dist-packages/torch/utils/data/dataloader.py in _next_data(self)
1201 else:
1202 del self._task_info[idx]
-> 1203 return self._process_data(data)
1204
1205 def _try_put_index(self):
/usr/local/lib/python3.7/dist-packages/torch/utils/data/dataloader.py in _process_data(self, data)
1227 self._try_put_index()
1228 if isinstance(data, ExceptionWrapper):
-> 1229 data.reraise()
1230 return data
1231
/usr/local/lib/python3.7/dist-packages/torch/_utils.py in reraise(self)
432 # instantiate since we don't know how to
433 raise RuntimeError(msg) from None
--> 434 raise exception
435
436
ValueError: Caught ValueError in DataLoader worker process 0.
Original Traceback (most recent call last):
File "/usr/local/lib/python3.7/dist-packages/torch/utils/data/_utils/worker.py", line 287, in _worker_loop
data = fetcher.fetch(index)
File "/usr/local/lib/python3.7/dist-packages/torch/utils/data/_utils/fetch.py", line 32, in fetch
data.append(next(self.dataset_iter))
File "/usr/local/lib/python3.7/dist-packages/detectron2/data/common.py", line 201, in __iter__
yield self.dataset[idx]
File "/usr/local/lib/python3.7/dist-packages/detectron2/data/common.py", line 90, in __getitem__
data = self._map_func(self._dataset[cur_idx])
File "/usr/local/lib/python3.7/dist-packages/detectron2/utils/serialize.py", line 26, in __call__
return self._obj(*args, **kwargs)
File "/usr/local/lib/python3.7/dist-packages/detectron2/data/dataset_mapper.py", line 189, in __call__
self._transform_annotations(dataset_dict, transforms, image_shape)
File "/usr/local/lib/python3.7/dist-packages/detectron2/data/dataset_mapper.py", line 128, in _transform_annotations
for obj in dataset_dict.pop("annotations")
File "/usr/local/lib/python3.7/dist-packages/detectron2/data/dataset_mapper.py", line 129, in <listcomp>
if obj.get("iscrowd", 0) == 0
File "/usr/local/lib/python3.7/dist-packages/detectron2/data/detection_utils.py", line 314, in transform_instance_annotations
annotation["keypoints"], transforms, image_size, keypoint_hflip_indices
File "/usr/local/lib/python3.7/dist-packages/detectron2/data/detection_utils.py", line 360, in transform_keypoint_annotations
"contains {} points!".format(len(keypoints), len(keypoint_hflip_indices))
ValueError: Keypoint data has 1 points, but metadata contains 18 points!
我检查了其他人的 Colab 和 Git，但他们在加载在单个类别中注释的关键点方面似乎没有问题，就像我对数据集所做的那样。
如果您对解决我在培训过程中遇到的这个问题有任何建议，请随时在这里与我分享您的一些知识......谢谢
基本上，我在这里犯的错误是
我给出了 18（因为我认为探测器是从 1 开始计数），而不是 17（0 到 17）。
吸取的教训...正确计数。
cfg.MODEL.ROI_KEYPOINT_HEAD.NUM_KEYPOINTS = 18
cfg.TEST.KEYPOINT_OKS_SIGMAS = np.ones((18, 1), dtype=float).tolist()