Hugging Face's logo

Datasets:
xiaomoguhzz
/
DeCLIP-TPAMI-data-2026-01

Dataset card Files
xet
 Community
1
Dataset Viewer
Duplicate
The dataset viewer is not available for this split.
Cannot load the dataset split (in streaming mode) to extract the first rows.
Error code:   StreamingRowsError
Exception:    CastError
Message:      Couldn't cast
models: list<item: string>
  child 0, item: string
mPC: struct<ClearCLIP: struct<base_ap50: double, novel_ap50: double, all_ap50: double>, CLIPSelf: struct< (... 57 chars omitted)
  child 0, ClearCLIP: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
      child 0, base_ap50: double
      child 1, novel_ap50: double
      child 2, all_ap50: double
  child 1, CLIPSelf: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
      child 0, base_ap50: double
      child 1, novel_ap50: double
      child 2, all_ap50: double
category_mPC: struct<ClearCLIP: struct<noise: struct<base_ap50: double, novel_ap50: double, all_ap50: double>, blu (... 525 chars omitted)
  child 0, ClearCLIP: struct<noise: struct<base_ap50: double, novel_ap50: double, all_ap50: double>, blur: struct<base_ap5 (... 197 chars omitted)
      child 0, noise: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
          child 0, base_ap50: double
          child 1, novel_ap50: double
          child 2, all_ap50: double
      child 1, blur: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
          child 0, base_ap50: double
          child 1, novel_ap50: double
          child 2, all_ap50: double
      child 2, weather: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
          child 0, base_ap50: double
          child 1, novel_ap50: double
          child 2, all_ap50: double
      child 3, digital: struct<base_ap50: double, novel_ap50: double, all
...
ouble
              child 1, novel_ap50: double
              child 2, all_ap50: double
          child 4, 5: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
              child 0, base_ap50: double
              child 1, novel_ap50: double
              child 2, all_ap50: double
      child 14, jpeg_compression: struct<1: struct<base_ap50: double, novel_ap50: double, all_ap50: double>, 2: struct<base_ap50: doub (... 246 chars omitted)
          child 0, 1: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
              child 0, base_ap50: double
              child 1, novel_ap50: double
              child 2, all_ap50: double
          child 1, 2: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
              child 0, base_ap50: double
              child 1, novel_ap50: double
              child 2, all_ap50: double
          child 2, 3: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
              child 0, base_ap50: double
              child 1, novel_ap50: double
              child 2, all_ap50: double
          child 3, 4: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
              child 0, base_ap50: double
              child 1, novel_ap50: double
              child 2, all_ap50: double
          child 4, 5: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
              child 0, base_ap50: double
              child 1, novel_ap50: double
              child 2, all_ap50: double
to
{'metrics': {'gaussian_noise': {'1': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '2': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '3': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '4': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '5': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}}, 'shot_noise': {'1': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '2': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '3': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '4': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '5': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}}, 'impulse_noise': {'1': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '2': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '3': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '4': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '5': {'base_ap50': Value('float64
...
novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '2': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '3': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '4': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '5': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}}, 'pixelate': {'1': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '2': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '3': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '4': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '5': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}}, 'jpeg_compression': {'1': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '2': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '3': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '4': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '5': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}}}}}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 99, in get_rows_or_raise
                  return get_rows(
                         ^^^^^^^^^
                File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
                  return func(*args, **kwargs)
                         ^^^^^^^^^^^^^^^^^^^^^
                File "/src/services/worker/src/worker/utils.py", line 77, in get_rows
                  rows_plus_one = list(itertools.islice(ds, rows_max_number + 1))
                                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2690, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2227, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2251, in _iter_arrow
                  for key, pa_table in self.ex_iterable._iter_arrow():
                                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 494, in _iter_arrow
                  for key, pa_table in iterator:
                                       ^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 384, in _iter_arrow
                  for key, pa_table in self.generate_tables_fn(**gen_kwags):
                                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 289, in _generate_tables
                  self._cast_table(pa_table, json_field_paths=json_field_paths),
                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 124, in _cast_table
                  pa_table = table_cast(pa_table, self.info.features.arrow_schema)
                             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2272, in table_cast
                  return cast_table_to_schema(table, schema)
                         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2218, in cast_table_to_schema
                  raise CastError(
              datasets.table.CastError: Couldn't cast
              models: list<item: string>
                child 0, item: string
              mPC: struct<ClearCLIP: struct<base_ap50: double, novel_ap50: double, all_ap50: double>, CLIPSelf: struct< (... 57 chars omitted)
                child 0, ClearCLIP: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
                    child 0, base_ap50: double
                    child 1, novel_ap50: double
                    child 2, all_ap50: double
                child 1, CLIPSelf: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
                    child 0, base_ap50: double
                    child 1, novel_ap50: double
                    child 2, all_ap50: double
              category_mPC: struct<ClearCLIP: struct<noise: struct<base_ap50: double, novel_ap50: double, all_ap50: double>, blu (... 525 chars omitted)
                child 0, ClearCLIP: struct<noise: struct<base_ap50: double, novel_ap50: double, all_ap50: double>, blur: struct<base_ap5 (... 197 chars omitted)
                    child 0, noise: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
                        child 0, base_ap50: double
                        child 1, novel_ap50: double
                        child 2, all_ap50: double
                    child 1, blur: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
                        child 0, base_ap50: double
                        child 1, novel_ap50: double
                        child 2, all_ap50: double
                    child 2, weather: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
                        child 0, base_ap50: double
                        child 1, novel_ap50: double
                        child 2, all_ap50: double
                    child 3, digital: struct<base_ap50: double, novel_ap50: double, all
              ...
              ouble
                            child 1, novel_ap50: double
                            child 2, all_ap50: double
                        child 4, 5: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
                            child 0, base_ap50: double
                            child 1, novel_ap50: double
                            child 2, all_ap50: double
                    child 14, jpeg_compression: struct<1: struct<base_ap50: double, novel_ap50: double, all_ap50: double>, 2: struct<base_ap50: doub (... 246 chars omitted)
                        child 0, 1: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
                            child 0, base_ap50: double
                            child 1, novel_ap50: double
                            child 2, all_ap50: double
                        child 1, 2: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
                            child 0, base_ap50: double
                            child 1, novel_ap50: double
                            child 2, all_ap50: double
                        child 2, 3: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
                            child 0, base_ap50: double
                            child 1, novel_ap50: double
                            child 2, all_ap50: double
                        child 3, 4: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
                            child 0, base_ap50: double
                            child 1, novel_ap50: double
                            child 2, all_ap50: double
                        child 4, 5: struct<base_ap50: double, novel_ap50: double, all_ap50: double>
                            child 0, base_ap50: double
                            child 1, novel_ap50: double
                            child 2, all_ap50: double
              to
              {'metrics': {'gaussian_noise': {'1': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '2': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '3': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '4': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '5': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}}, 'shot_noise': {'1': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '2': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '3': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '4': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '5': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}}, 'impulse_noise': {'1': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '2': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '3': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '4': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '5': {'base_ap50': Value('float64
              ...
              novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '2': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '3': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '4': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '5': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}}, 'pixelate': {'1': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '2': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '3': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '4': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '5': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}}, 'jpeg_compression': {'1': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '2': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '3': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '4': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}, '5': {'base_ap50': Value('float64'), 'novel_ap50': Value('float64'), 'all_ap50': Value('float64')}}}}}
              because column names don't match

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

DeCLIP 数据集标注文件

本目录包含 DeCLIP 项目使用的 COCO 数据集标注文件（仅标注，不含图片）。

📦 文件列表

核心标注 (coco_annotations_core.zip)

  • instances_train2017.json - COCO 实例分割标注（训练集）
  • instances_val2017.json - COCO 实例分割标注（验证集）
  • panoptic_train2017.json - COCO 全景分割标注（训练集）
  • panoptic_val2017.json - COCO 全景分割标注（验证集）
  • captions_train2017_tags_allcaps.json - COCO 描述标注

全景分割 PNG masks (coco_panoptic_masks.zip)

  • panoptic_train2017/ - 训练集全景分割 PNG masks (118,287 个文件)
  • panoptic_val2017/ - 验证集全景分割 PNG masks (5,000 个文件)

转换后标注 (coco_converted_annotations.zip)

  • png_coco_train2017.json - 转换为 PNG 格式的训练集标注
  • png_coco_val2017.json - 转换为 PNG 格式的验证集标注
  • train2017/ - labelTrainIds 格式的训练集 PNG masks
  • val2017/ - labelTrainIds 格式的验证集 PNG masks

LVIS 标注 (lvis_annotations.zip) - 可选

  • lvis_v1_train.json - LVIS v1.0 训练标注 (1203 类，基于 COCO 图片)

RefCOCO 系列标注 (refcoco_annotations.zip) - 可选

  • refcoco/ - RefCOCO 数据集标注
  • refcoco+/ - RefCOCO+ 数据集标注
  • refcocog/ - RefCOCOg 数据集标注

🚀 使用方法

1. 下载标注文件 

# 下载所有 DeCLIP 标注
huggingface-cli download xiaomoguhzz/xiaomogu_pami_dataset \
  --include "declip_annotations/*" \
  --repo-type dataset \
  --local-dir ./declip_data

# 或单独下载某个文件
huggingface-cli download xiaomoguhzz/xiaomogu_pami_dataset \
  declip_annotations/coco_annotations_core.zip \
  --repo-type dataset \
  --local-dir ./declip_data

2. 解压标注 

cd declip_data/declip_annotations
unzip coco_annotations_core.zip -d ../../datasets/coco/annotations/
unzip coco_panoptic_masks.zip -d ../../datasets/coco/annotations/
unzip coco_converted_annotations.zip -d ../../datasets/coco/annotations/
unzip lvis_annotations.zip -d ../../datasets/coco/annotations/
unzip refcoco_annotations.zip -d ../../datasets/coco/annotations/

3. 下载 COCO 图片（单独下载） 

cd datasets/coco/

# 训练集图片 (19GB)
wget http://images.cocodataset.org/zips/train2017.zip
unzip train2017.zip

# 验证集图片 (1GB)
wget http://images.cocodataset.org/zips/val2017.zip
unzip val2017.zip

💾 存储空间需求

项目 大小 说明
标注文件（压缩） ~2.75 GB 5 个 zip 文件
标注文件（解压） ~3-4 GB 解压后
COCO 图片 ~20 GB 需单独下载
总计 ~23-24 GB 完整数据集

📚 相关链接

📄 许可证

Downloads last month
144
Total file size:
13 GB