Evaluation Dataset

The following datasets are all transformed into standard Evaluation Prompts before evaluation.

Dataset 1（RefCOCO）

#Evaluation Metrics-1. Precision or Accuracy

Data description：

RefCOCO: an referring expression/visual grounding dataset, with images fromMS COCO Website, whose image and corresponding phrases(referring expressions) are manually annotated. RefCOCO is a small sized vision language multimodal training and testing benchmark dataset, which can be used to evaluate tasks such as referring expression/visual grounding. The image content mostly comes from real-life scenarios, and the phrases(referring expressions) descriptions are usually intuitive descriptions of the image content.

Amount of source data：

The dataset is split into train(120624), validation(10834), testA(5657), testB(5095)

Amount of evaluation data：

The evaluation data are instances of 5657 image regions and corresponding phrases(referring expressions) from testA split and 5095 image regions and corresponding phrases(referring expressions) from testB split.

Data detail：

KEYS	EXPLAIN
'images'
id	image id
file_name	image filename
width	image width
height	image height
coco_url	image url in coco dataset
flickr_url	image url in flickr dataset
license	image license type
'annotations'
id	image region id
image_id	image region's image id
category_id	image region's category id
bbox	image region's bounding box
segmentation	image region's polygon edge segmentation
area	image region's segmentation area
iscrowd	image region's segmentation is multi-object crowd
'references'
ref_id	referring expression id
ann_id	referring expression's image region id
split	referring expression's dataset split
sent_ids	referring expressions' indices list
sentences	referring expressions' sentence contents

Sample of source dataset：

import json as jsonmod import pickle refcoco = jsonmod.load(open('./refcoco/instances.json', 'r')) refcoco_p = pickle.load(open('./refcoco/refs(unc).p', 'rb'),fix_imports=True)

refcoco_p[0]
{'sent_ids': [0, 1, 2], 
 'file_name': 'COCO_train2014_000000581857_16.jpg', 
 'ann_id': 1719310, 
 'ref_id': 0, 
 'image_id': 581857, 
 'split': 'train', 
 'sentences': 
    [{'tokens': ['the', 'lady', 'with', 'the', 'blue', 'shirt'], 'raw': 'THE LADY WITH THE BLUE SHIRT', 'sent_id': 0, 'sent': 'the lady with the blue shirt'}, 
     {'tokens': ['lady', 'with', 'back', 'to', 'us'], 'raw': 'lady w back to us', 'sent_id': 1, 'sent': 'lady with back to us'}, 
     {'tokens': ['blue', 'shirt'], 'raw': 'blue shirt', 'sent_id': 2, 'sent': 'blue shirt'}
    ], 
 'category_id': 1
}
refcoco_p[50000-1]
{'sent_ids': [142208, 142209], 
 'file_name': 'COCO_train2014_000000000072_0.jpg', 
 'ann_id': 598731, 
 'ref_id': 49999, 
 'image_id': 72, 
 'split': 'train', 
 'sentences': 
    [{'tokens': ['right', 'giraffe'], 'raw': 'RIGHT GIRAFFE', 'sent_id': 142208, 'sent': 'right giraffe'}, 
     {'tokens': ['right', 'girafe'], 'raw': 'right girafe', 'sent_id': 142209, 'sent': 'right girafe'}
    ], 
 'category_id': 25
}

refcoco['images'][0]
{'license': 1, 
 'file_name': 'COCO_train2014_000000098304.jpg', 
 'coco_url': 'http://mscoco.org/images/98304', 
 'height': 424, 
 'width': 640, 
 'date_captured': '2013-11-21 23:06:41', 
 'flickr_url': 'http://farm6.staticflickr.com/5062/5896644212_a326e96ea9_z.jpg', 
 'id': 98304
}
refcoco['images'][19994-1]
{'license': 6, 
 'file_name': 'COCO_train2014_000000458751.jpg', 
 'coco_url': 'http://mscoco.org/images/458751', 
 'height': 576, 
 'width': 592, 
 'date_captured': '2013-11-16 21:13:51', 
 'flickr_url': 'http://farm8.staticflickr.com/7018/6821165845_48ebd9590f_z.jpg', 
 'id': 458751
}

refcoco['annotations'][0]
{'segmentation': [[267.52, 229.75, 265.6, 226.68, 265.79, 223.6, 263.87, 220.15, 263.87, 216.88, 266.94, 217.07, 268.48, 221.3, 272.32, 219.95, 276.35, 220.15, 279.62, 218.03, 283.46, 218.42, 285.0, 220.92, 285.0, 223.22, 284.42, 224.95, 280.96, 225.14, 279.81, 226.48, 281.73, 228.41, 279.43, 229.37, 275.78, 229.17, 273.86, 229.56, 274.24, 232.05, 269.82, 231.67, 267.14, 231.48, 266.75, 228.6]], 
 'area': 197.29899999999986, 
 'iscrowd': 0, 
 'image_id': 98304, 
 'bbox': [263.87, 216.88, 21.13, 15.17], 
 'category_id': 18, 
 'id': 3007
}
refcoco['annotations'][196771-1]
{'segmentation': [[203.42, 96.23, 216.68, 104.44, 216.05, 114.54, 226.15, 118.96, 228.67, 132.21, 247.61, 138.52, 250.13, 156.83, 236.88, 159.35, 234.35, 167.56, 274.12, 168.19, 281.69, 185.87, 284.85, 213.01, 267.81, 237.62, 243.19, 236.36, 238.14, 223.74, 232.46, 232.57, 231.2, 284.33, 159.87, 283.07, 159.87, 218.06, 151.67, 206.7, 154.19, 190.92, 159.87, 184.6, 158.61, 166.3, 140.3, 153.04, 142.2, 144.84, 178.81, 147.99, 183.86, 142.94, 169.97, 125.9, 173.13, 114.54, 176.28, 113.91, 185.75, 96.87, 200.9, 94.97]], 
 'area': 16238.20485, 
 'iscrowd': 0, 
 'image_id': 458751, 
 'bbox': [140.3, 94.97, 144.55, 189.36], 
 'category_id': 11, 
 'id': 1808941
}

Dataset structure：

Dataset 2（RefCOCO+）

#Evaluation Metrics-1. Precision or Accuracy

Data description：

RefCOCO+: an referring expression/visual grounding dataset, with images fromMS COCO Website, whose image and corresponding phrases(referring expressions) are manually annotated. RefCOCO+ is a small sized vision language multimodal training and testing benchmark dataset, which can be used to evaluate tasks such as referring expression/visual grounding. The image content mostly comes from real-life scenarios, and the phrases(referring expressions) descriptions are usually intuitive descriptions of the image content.

Amount of source data：

The dataset is split into train(120191), validation(10758), testA(5726), testB(4889)

Amount of evaluation data：

The evaluation data are instances of 5726 image regions and corresponding phrases(referring expressions) from testA split and 4889 image regions and corresponding phrases(referring expressions) from testB split.

Data detail：

KEYS	EXPLAIN
'images'
id	image id
file_name	image filename
width	image width
height	image height
coco_url	image url in coco dataset
flickr_url	image url in flickr dataset
license	image license type
'annotations'
id	image region id
image_id	image region's image id
category_id	image region's category id
bbox	image region's bounding box
segmentation	image region's polygon edge segmentation
area	image region's segmentation area
iscrowd	image region's segmentation is multi-object crowd
'references'
ref_id	referring expression id
ann_id	referring expression's image region id
split	referring expression's dataset split
sent_ids	referring expressions' indices list
sentences	referring expressions' sentence contents

Sample of source dataset：

import json as jsonmod refcoco_plus = jsonmod.load(open('./refcoco+/instances.json', 'r')) refcoco_plus_p = pickle.load(open('./refcoco+/refs(unc).p', 'rb'),fix_imports=True)

refcoco_plus_p[0]
{'sent_ids': [0, 1, 2], 
 'file_name': 'COCO_train2014_000000581857_16.jpg', 
 'ann_id': 1719310, 
 'ref_id': 0, 
 'image_id': 581857, 
 'split': 'train', 
 'sentences': [
    {'tokens': ['navy', 'blue', 'shirt'], 'raw': 'navy blue shirt', 'sent_id': 0, 'sent': 'navy blue shirt'}, 
    {'tokens': ['woman', 'back', 'in', 'blue'], 'raw': 'woman back in blue', 'sent_id': 1, 'sent': 'woman back in blue'}, 
    {'tokens': ['blue', 'shirt'], 'raw': 'blue shirt', 'sent_id': 2, 'sent': 'blue shirt'}
 ], 
 'category_id': 1
}
refcoco_plus_p[49856-1]
{'sent_ids': [141560, 141561, 141562, 141563], 
 'file_name': 'COCO_train2014_000000000072_0.jpg', 
 'ann_id': 598731, 
 'ref_id': 49855, 
 'image_id': 72, 
 'split': 'train', 
 'sentences': [
    {'tokens': ['shorter', 'giraffe'], 'raw': 'shorter giraffe', 'sent_id': 141560, 'sent': 'shorter giraffe'}, 
    {'tokens': ['giraffe', 'closest', 'to', 'camera'], 'raw': 'giraffe closest to camera', 'sent_id': 141561, 'sent': 'giraffe closest to camera'}, 
    {'tokens': ['bent', 'neck'], 'raw': 'bent neck', 'sent_id': 141562, 'sent': 'bent neck'}, 
    {'tokens': ['shorter', 'animal'], 'raw': 'shorter animal', 'sent_id': 141563, 'sent': 'shorter animal'}
 ], 
 'category_id': 25
}

refcoco_plus['images'][0]
{'license': 1, 
 'file_name': 'COCO_train2014_000000098304.jpg', 
 'coco_url': 'http://mscoco.org/images/98304', 
 'height': 424, 
 'width': 640, 
 'date_captured': '2013-11-21 23:06:41', 
 'flickr_url': 'http://farm6.staticflickr.com/5062/5896644212_a326e96ea9_z.jpg', 
 'id': 98304
}
refcoco_plus['images'][19992-1]
{'license': 6, 
 'file_name': 'COCO_train2014_000000458751.jpg', 
 'coco_url': 'http://mscoco.org/images/458751', 
 'height': 576, 
 'width': 592, 
 'date_captured': '2013-11-16 21:13:51', 
 'flickr_url': 'http://farm8.staticflickr.com/7018/6821165845_48ebd9590f_z.jpg', 
 'id': 458751
}

refcoco_plus['annotations'][0]
{'segmentation': [[267.52, 229.75, 265.6, 226.68, 265.79, 223.6, 263.87, 220.15, 263.87, 216.88, 266.94, 217.07, 268.48, 221.3, 272.32, 219.95, 276.35, 220.15, 279.62, 218.03, 283.46, 218.42, 285.0, 220.92, 285.0, 223.22, 284.42, 224.95, 280.96, 225.14, 279.81, 226.48, 281.73, 228.41, 279.43, 229.37, 275.78, 229.17, 273.86, 229.56, 274.24, 232.05, 269.82, 231.67, 267.14, 231.48, 266.75, 228.6]], 
 'area': 197.29899999999986, 
 'iscrowd': 0, 
 'image_id': 98304, 
 'bbox': [263.87, 216.88, 21.13, 15.17], 
 'category_id': 18, 
 'id': 3007
}
refcoco_plus['annotations'][196737-1]
{'segmentation': [[203.42, 96.23, 216.68, 104.44, 216.05, 114.54, 226.15, 118.96, 228.67, 132.21, 247.61, 138.52, 250.13, 156.83, 236.88, 159.35, 234.35, 167.56, 274.12, 168.19, 281.69, 185.87, 284.85, 213.01, 267.81, 237.62, 243.19, 236.36, 238.14, 223.74, 232.46, 232.57, 231.2, 284.33, 159.87, 283.07, 159.87, 218.06, 151.67, 206.7, 154.19, 190.92, 159.87, 184.6, 158.61, 166.3, 140.3, 153.04, 142.2, 144.84, 178.81, 147.99, 183.86, 142.94, 169.97, 125.9, 173.13, 114.54, 176.28, 113.91, 185.75, 96.87, 200.9, 94.97]], 
 'area': 16238.20485, 
 'iscrowd': 0, 
 'image_id': 458751, 
 'bbox': [140.3, 94.97, 144.55, 189.36], 
 'category_id': 11, 
 'id': 1808941
}

Dataset structure：

Dataset 3（RefCOCOg）

#Evaluation Metrics-1. Precision or Accuracy

Data description：

RefCOCOg: an referring expression/visual grounding dataset, with images fromMS COCO Website, whose image and corresponding phrases(referring expressions) are manually annotated. RefCOCOg is a small sized vision language multimodal training and testing benchmark dataset, which can be used to evaluate tasks such as referring expression/visual grounding. The image content mostly comes from real-life scenarios, and the phrases(referring expressions) descriptions are usually intuitive descriptions of the image content.

Amount of source data：

The dataset is split into train(80512), validation(4896), test(9602)

Amount of evaluation data：

The evaluation data are instances of 9602 image regions and corresponding phrases(referring expressions) from test split.

Data detail：

KEYS	EXPLAIN
'images'
id	image id
file_name	image filename
width	image width
height	image height
coco_url	image url in coco dataset
flickr_url	image url in flickr dataset
license	image license type
'annotations'
id	image region id
image_id	image region's image id
category_id	image region's category id
bbox	image region's bounding box
segmentation	image region's polygon edge segmentation
area	image region's segmentation area
iscrowd	image region's segmentation is multi-object crowd
'references'
ref_id	referring expression id
ann_id	referring expression's image region id
split	referring expression's dataset split
sent_ids	referring expressions' indices list
sentences	referring expressions' sentence contents

Sample of source dataset：

import json as jsonmod refcoco_g = jsonmod.load(open('./refcocog/instances.json', 'r')) refcoco_g_p = pickle.load(open('./refcocog/refs(umd).p', 'rb'),fix_imports=True)

refcoco_g_p[0]
{'image_id': 380440, 
 'split': 'test', 
 'sentences': [
    {'tokens': ['the', 'man', 'in', 'yellow', 'coat'], 'raw': 'the man in yellow coat', 'sent_id': 8, 'sent': 'the man in yellow coat'}, 
    {'tokens': ['skiier', 'in', 'red', 'pants'], 'raw': 'Skiier in red pants.', 'sent_id': 9, 'sent': 'skiier in red pants'}
 ], 
 'file_name': 'COCO_train2014_000000380440_491042.jpg', 
 'category_id': 1, 
 'ann_id': 491042, 
 'sent_ids': [8, 9], 
 'ref_id': 0
}
refcoco_g_p[49822-1]
{'image_id': 573297, 
 'split': 'train', 
 'sentences': [
    {'tokens': ['a', 'person', 'in', 'red', 'dress', 'and', 'he', 'is', 'seeing', 'his', 'mobile'], 'raw': 'A person in red dress and he is seeing his mobile.', 'sent_id': 104558, 'sent': 'a person in red dress and he is seeing his mobile'}, 
    {'tokens': ['man', 'wearing', 'a', 'red', 'costume'], 'raw': 'Man wearing a red costume.', 'sent_id': 104559, 'sent': 'man wearing a red costume'}
 ], 
 'file_name': 'COCO_train2014_000000573297_472971.jpg', 
 'category_id': 1, 
 'ann_id': 472971, 
 'sent_ids': [104558, 104559], 
 'ref_id': 49821
}

refcoco_g['images'][0]
{'license': 1, 
 'file_name': 'COCO_train2014_000000131074.jpg', 
 'coco_url': 'http://mscoco.org/images/131074', 
 'height': 428, 
 'width': 640, 
 'date_captured': '2013-11-21 01:03:06', 
 'flickr_url': 'http://farm9.staticflickr.com/8308/7908210548_33e532d119_z.jpg', 
 'id': 131074
}
refcoco_g['images'][25799-1]
{'license': 5, 
 'file_name': 'COCO_train2014_000000524286.jpg', 
 'coco_url': 'http://mscoco.org/images/524286', 
 'height': 480, 
 'width': 640, 
 'date_captured': '2013-11-22 01:08:02', 
 'flickr_url': 'http://farm4.staticflickr.com/3286/3160643026_c2691d2c55_z.jpg', 
 'id': 524286
}

refcoco_g['annotations'][0]
{'segmentation': [[21.11, 239.09, 16.31, 274.6, 198.65, 349.45, 240.87, 336.98, 320.52, 293.79, 334.91, 248.69, 357.95, 273.64, 353.15, 289.0, 398.25, 267.88, 437.6, 251.57, 412.65, 228.54, 240.87, 210.31, 219.76, 141.21, 113.24, 153.69, 63.34, 156.57, 26.87, 169.04]], 
 'area': 48667.84089999999, 
 'iscrowd': 0, 
 'image_id': 131074, 
 'bbox': [16.31, 141.21, 421.29, 208.24], 
 'category_id': 65, 
 'id': 318235
}
refcoco_g['annotations'][208960-1]
{'segmentation': [[158.56, 212.49, 158.56, 94.92, 467.06, 85.21, 476.76, 209.26]], 
 'area': 37887.193, 
 'iscrowd': 0, 
 'image_id': 524286, 
 'bbox': [158.56, 85.21, 318.2, 127.28], 
 'category_id': 76, 
 'id': 1635174
}

Dataset structure：

Citation information：

{RefCOCO, RefCOCO+,
  title={Modeling context in referring expressions},
  author={Yu, Licheng and Poirson, Patrick and Yang, Shan and Berg, Alexander C and Berg, Tamara L},
  booktitle={Computer Vision--ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11-14, 2016, Proceedings, Part II 14},
  pages={69--85},
  year={2016},
  organization={Springer}
}
{RefCOCOg,
  title={Generation and comprehension of unambiguous object descriptions},
  author={Mao, Junhua and Huang, Jonathan and Toshev, Alexander and Camburu, Oana and Yuille, Alan L and Murphy, Kevin},
  booktitle={Proceedings of the IEEE conference on computer vision and pattern recognition},
  pages={11--20},
  year={2016}
}

Licensing information：

[Datasets] ALL RefCOCO* Datasets MS COCO Image Dataset RefCOCO RefCOCO+RefCOCOg [Licenses] Attribution-NonCommercial-ShareAlike License Attribution-NonCommercial License Attribution-NonCommercial-NoDerivs License Attribution License Attribution-ShareAlike License Attribution-NoDerivs License No known copyright restrictions United States Government Work

Dataset 4 (ARPGrounding)

#Evaluation Metrics-1. Precision or Accuracy

Data description：

ARPGrounding is a visual grounding benchmark designed to evaluate the compositional reasoning ability of vision-language models (VLMs). The dataset is built on top of the Visual Genome (VG) dataset, constructed via dependency parsing and further manually filtered. ARPGrounding aims to assess fine-grained compositional understanding in vision and language, especially in confusing scenarios where current models are prone to failure. Images come from real, complex visual scenes. The referring phrases are organized as ambiguous positive/negative pairs, requiring the model to distinguish two visually similar or semantically related objects in the same image based on attribute, relation, or priority differences.

Amount of source data：

The dataset is derived from 108,249 images in Visual Genome and then filtered/constructed to form ARPGrounding.

Amount of evaluation data：

The evaluation set contains 11,425 region-level instances paired with phrases (referring expressions). It consists of three subsets:

Attribute subset: 6,632 samples
Relation subset: 370 samples
Priority subset: 4,423 samples

Data detail：

KEYS	EXPLAIN
`attribute_vg.pkl`	Attribute subset
phrase	Positive/negative phrase describing an object (e.g., `red building`)
x	Top-left x coordinate of the bounding box
y	Top-left y coordinate of the bounding box
w	Bounding box width
h	Bounding box height
attributes	List of attributes (e.g., `['brown', 'red']`)
names	List of object names (e.g., `['building']`)
object_id	Unique object ID in Visual Genome
synsets	WordNet synsets (e.g., `['building.n.01']`)
`relationship_vg.pkl`	Relation subset
phrase	Phrase describing spatial/action relations (e.g., `grass on top of sand`)
x, y, w, h	Bounding box coordinates of the involved region (xywh format)
`priority_vg.pkl`	Priority/Saliency subset
phrase	Phrase describing position priority or contrastive relations
x, y, w, h	Bounding box coordinates of the salient region (xywh format)
Data hierarchy
num_images	Total number of images in the dataset
pairs	List of positive/negative paired samples for each image
pair[0] (pos)	Positive sample dictionary
pair[1] (neg)	Negative sample dictionary

Sample of source dataset：

import pickle

data = pickle.load(open(attribute_vg.pkl, "rb"))

Example 1: Attribute - Disambiguating Color

Example (Attribute VG Pairs, positive and negative):

    [
        {
            'synsets': ['building.n.01'],
            'h': 298,
            'object_id': 1023846,
            'names': ['building'],
            'w': 282,
            'attributes': ['brown', 'red'],
            'y': 13,
            'x': 165,
            'phrase': 'red building'
        },
        {
            'synsets': ['building.n.01'],
            'h': 384,
            'object_id': 1023819,
            'names': ['building'],
            'w': 251,
            'attributes': ['orange', 'brown', 'tall'],
            'y': 4,
            'x': 547,
            'phrase': 'orange building'
        }
    ]

Example 2: Relation - Disambiguating Spatial Position

Example (Relation VG Pairs, positive and negative):

    [
        {
            'h': 156,
            'w': 799,
            'y': 264,
            'x': 1,
            'phrase': 'grass on top of sand'
        },
        {
            'h': 80,
            'w': 374,
            'y': 468,
            'x': 273,
            'phrase': 'grass in sand'
        }
    ]

Example 3: Priority - Disambiguating the Subject

Example (Priority VG Pairs, positive and negative):

    [
        {
            'h': 200,
            'w': 196,
            'y': 391,
            'x': 294,
            'phrase': 'cpu on floor'
        },
        {
            'h': 73,
            'w': 402,
            'y': 526,
            'x': 237,
            'phrase': 'floor under cpu'
        }
    ]

Dataset structure：

Dataset Composition and Protocol

ARPGrounding consists of three core parts, aiming to test compositional reasoning along different dimensions:

Attribute: 6,632 samples. Tests whether a model can distinguish objects of the same category with different attributes (e.g., color/material/state). Example: distinguishing “a brown dog” vs “a black dog”.
Relation: 370 samples. Tests whether a model can identify the target purely based on relations between objects (typically spatial/action relations). Example: “a computer on the table” vs “a computer under the table”.
Priority: 4,423 samples. Tests whether a model can correctly identify the grammatical subject, without being misled by other nouns mentioned in the text. Often involves complex sentence structures where subject/object positions are swapped.

Citation information：

    @inproceedings{ARPGrounding,
    title={Investigating Compositional Challenges in Vision-Language Models for Visual Grounding},
    author={Zeng, Yunan and Huang, Yan and Zhang, Jinjin and Jie, Zequn and Chai, Zhenhua and Wang, Liang},
    booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
    pages={14141--14151},
    year={2024}
    }

Source Dataset / Copyright & Usage Notes

[Datasets] ARPGrounding (Based on Visual Genome)

[Licenses] This dataset is derived from Visual Genome. Image licenses follow the original Visual Genome image sources (e.g., Flickr) and may vary per image (e.g., Attribution, NonCommercial, NoDerivs, ShareAlike, etc.). Please refer to the Visual Genome website for licensing information and use the images according to their original licenses. ARPGrounding annotations and processing scripts are provided under the same license as this repository unless otherwise stated.

Evaluation Dataset ​

Dataset 1（RefCOCO） ​

Data description： ​

Amount of source data： ​

Amount of evaluation data： ​

Data detail： ​

Sample of source dataset： ​

Dataset structure： ​

Dataset 2（RefCOCO+） ​

Data description： ​

Amount of source data： ​

Amount of evaluation data： ​

Data detail： ​

Sample of source dataset： ​

Dataset structure： ​

Dataset 3（RefCOCOg） ​

Data description： ​

Amount of source data： ​

Amount of evaluation data： ​

Data detail： ​

Sample of source dataset： ​

Dataset structure： ​

Citation information： ​

Licensing information： ​

Dataset 4 (ARPGrounding) ​

Data description： ​

Amount of source data： ​

Amount of evaluation data： ​

Data detail： ​

Sample of source dataset： ​

Example 1: Attribute - Disambiguating Color ​

Example 2: Relation - Disambiguating Spatial Position ​

Example 3: Priority - Disambiguating the Subject ​

Dataset structure： ​

Citation information： ​

Source Dataset / Copyright & Usage Notes ​

Evaluation Dataset

Dataset 1（RefCOCO）

Data description：

Amount of source data：

Amount of evaluation data：

Data detail：

Sample of source dataset：

Dataset structure：

Dataset 2（RefCOCO+）

Data description：

Amount of source data：

Amount of evaluation data：

Data detail：

Sample of source dataset：

Dataset structure：

Dataset 3（RefCOCOg）

Data description：

Amount of source data：

Amount of evaluation data：

Data detail：

Sample of source dataset：

Dataset structure：

Citation information：

Licensing information：

Dataset 4 (ARPGrounding)

Data description：

Amount of source data：

Amount of evaluation data：

Data detail：

Sample of source dataset：

Example 1: Attribute - Disambiguating Color

Example 2: Relation - Disambiguating Spatial Position

Example 3: Priority - Disambiguating the Subject

Dataset structure：

Citation information：

Source Dataset / Copyright & Usage Notes