MMdetection using coco dataset

[shabbbb]

Sir, During the training of mmdetection using coco dataset, I am getting the following error :

loading annotations into memory...

JSONDecodeError Traceback (most recent call last) /usr/local/lib/python3.7/dist-packages/mmcv/utils/registry.py in build_from_cfg(cfg, registry, default_args) 50 try: ---> 51 return obj_cls(**args) 52 except Exception as e:

10 frames /content/mmdetection/mmdet/datasets/custom.py in init(self, ann_file, pipeline, classes, data_root, img_prefix, seg_prefix, proposal_file, test_mode, filter_empty_gt) 87 # load annotations (and proposals) ---> 88 self.data_infos = self.load_annotations(self.ann_file) 89

in load_annotations(self, ann_file) 46 ---> 47 self.coco = COCO(ann_file) 48 # The order of returned `cat_ids` will not /content/mmdetection/mmdet/datasets/api_wrappers/coco_api.py in __init__(self, annotation_file) 21 UserWarning) ---> 22 super().__init__(annotation_file=annotation_file) 23 self.img_ann_map = self.imgToAnns /usr/local/lib/python3.7/dist-packages/pycocotools/coco.py in __init__(self, annotation_file) 84 with open(annotation_file, 'r') as f: ---> 85 dataset = json.load(f) 86 assert type(dataset)==dict, 'annotation file format {} not supported'.format(type(dataset)) /usr/lib/python3.7/json/__init__.py in load(fp, cls, object_hook, parse_float, parse_int, parse_constant, object_pairs_hook, **kw) 295 parse_float=parse_float, parse_int=parse_int, --> 296 parse_constant=parse_constant, object_pairs_hook=object_pairs_hook, **kw) 297 /usr/lib/python3.7/json/__init__.py in loads(s, encoding, cls, object_hook, parse_float, parse_int, parse_constant, object_pairs_hook, **kw) 347 parse_constant is None and object_pairs_hook is None and not kw): --> 348 return _default_decoder.decode(s) 349 if cls is None: /usr/lib/python3.7/json/decoder.py in decode(self, s, _w) 336 """ --> 337 obj, end = self.raw_decode(s, idx=_w(s, 0).end()) 338 end = _w(s, end).end() /usr/lib/python3.7/json/decoder.py in raw_decode(self, s, idx) 352 try: --> 353 obj, end = self.scan_once(s, idx) 354 except StopIteration as err: JSONDecodeError: Expecting property name enclosed in double quotes: line 1853893 column 12 (char 51380224) During handling of the above exception, another exception occurred: TypeError Traceback (most recent call last) in () 5 6 # Build dataset ----> 7 datasets = [build_dataset(cfg.data.train)] 8 9 # Build the detector /content/mmdetection/mmdet/datasets/builder.py in build_dataset(cfg, default_args) 69 dataset = _concat_dataset(cfg, default_args) 70 else: ---> 71 dataset = build_from_cfg(cfg, DATASETS, default_args) 72 73 return dataset /usr/local/lib/python3.7/dist-packages/mmcv/utils/registry.py in build_from_cfg(cfg, registry, default_args) 52 except Exception as e: 53 # Normal TypeError does not print class name. ---> 54 raise type(e)(f'{obj_cls.__name__}: {e}') 55 56 TypeError: __init__() missing 2 required positional arguments: 'doc' and 'pos' ----------------------------------------------------------------------------- Please help me to fix this. Thank you...

[BIGWangYuDong]

Hi, did you use your own dataset? or use coco original dataset? Also, please re-paste your error, it's hard to read. Your can use ```shell your error ``` which is seems like

your error 

[shabbbb]

Hey sir, I tried to use the original cocodataset, but the memory was running out. And so I used some part of the COCO dataset. But, the format is same as that of original cocodataset.The error is as follows:

loading annotations into memory...
---------------------------------------------------
JSONDecodeError                           Traceback (most recent call last)
/usr/local/lib/python3.7/dist-packages/mmcv/utils/registry.py in build_from_cfg(cfg, registry, default_args)
     50     try:
---> 51         return obj_cls(**args)
     52     except Exception as e:

10 frames
/content/mmdetection/mmdet/datasets/custom.py in __init__(self, ann_file, pipeline, classes, data_root, img_prefix, seg_prefix, proposal_file, test_mode, filter_empty_gt)
     87         # load annotations (and proposals)
---> 88         self.data_infos = self.load_annotations(self.ann_file)
     89 

<ipython-input-11-36b37ed1d5fb> in load_annotations(self, ann_file)
     46 
---> 47         self.coco = COCO(ann_file)
     48         # The order of returned `cat_ids` will not

/content/mmdetection/mmdet/datasets/api_wrappers/coco_api.py in __init__(self, annotation_file)
     21                 UserWarning)
---> 22         super().__init__(annotation_file=annotation_file)
     23         self.img_ann_map = self.imgToAnns

/usr/local/lib/python3.7/dist-packages/pycocotools/coco.py in __init__(self, annotation_file)
     84             with open(annotation_file, 'r') as f:
---> 85                 dataset = json.load(f)
     86             assert type(dataset)==dict, 'annotation file format {} not supported'.format(type(dataset))

/usr/lib/python3.7/json/__init__.py in load(fp, cls, object_hook, parse_float, parse_int, parse_constant, object_pairs_hook, **kw)
    295         parse_float=parse_float, parse_int=parse_int,
--> 296         parse_constant=parse_constant, object_pairs_hook=object_pairs_hook, **kw)
    297 

/usr/lib/python3.7/json/__init__.py in loads(s, encoding, cls, object_hook, parse_float, parse_int, parse_constant, object_pairs_hook, **kw)
    347             parse_constant is None and object_pairs_hook is None and not kw):
--> 348         return _default_decoder.decode(s)
    349     if cls is None:

/usr/lib/python3.7/json/decoder.py in decode(self, s, _w)
    336         """
--> 337         obj, end = self.raw_decode(s, idx=_w(s, 0).end())
    338         end = _w(s, end).end()

/usr/lib/python3.7/json/decoder.py in raw_decode(self, s, idx)
    352         try:
--> 353             obj, end = self.scan_once(s, idx)
    354         except StopIteration as err:

JSONDecodeError: Expecting property name enclosed in double quotes: line 1853893 column 12 (char 51380224)

During handling of the above exception, another exception occurred:

TypeError                                 Traceback (most recent call last)
<ipython-input-14-153948f490f2> in <module>()
      5 
      6 # Build dataset
----> 7 datasets = [build_dataset(cfg.data.train)]
      8 datasets

/content/mmdetection/mmdet/datasets/builder.py in build_dataset(cfg, default_args)
     69         dataset = _concat_dataset(cfg, default_args)
     70     else:
---> 71         dataset = build_from_cfg(cfg, DATASETS, default_args)
     72 
     73     return dataset

/usr/local/lib/python3.7/dist-packages/mmcv/utils/registry.py in build_from_cfg(cfg, registry, default_args)
     52     except Exception as e:
     53         # Normal TypeError does not print class name.
---> 54         raise type(e)(f'{obj_cls.__name__}: {e}')
     55 
     56 

TypeError: __init__() missing 2 required positional arguments: 'doc' and 'pos'

https://colab.research.google.com/drive/1JP1rZKYZwxYlKH0pPZst1WwaWn1OPmhK?usp=sharing

This is my code sir.. Thank you...

[WassimBouatay]

Did you solve your issue ? I have the same error can you help me ? I changed my dataset to the same form as it was stated in the tutorial

[shabbbb]

Hey, Is the dataset that you used were the original coco dataset? I guess, mmdetection only supports the original one. I was trying it in google colab. And found that colab does not provide enough memory to support the original coco dataset. So, I switched to pascal VC dataset. Thank you...

[PolarBear-J]

I had the same problem,but when I saw this discussion (https://github.com/ucbdrive/3d-vehicle-tracking/issues/21) ,I check my json file again.A small error appeared in it(an extra '],'),and then train.py is ok,although "CUDA out of memory"(^o^)