Datasets:

Modalities:
Tabular
Text
Formats:
parquet
Size:
< 1K
Libraries:
Datasets
pandas
License:
atwang commited on
Commit
3d92824
·
verified ·
1 Parent(s): d9833e0

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +145 -86
README.md CHANGED
@@ -1,86 +1,145 @@
1
- ---
2
- license: mit
3
- dataset_info:
4
- features:
5
- - name: scene_id
6
- dtype: string
7
- - name: object_name
8
- dtype: string
9
- - name: description
10
- dtype: string
11
- - name: num_attributes_target
12
- dtype: int64
13
- - name: num_attributes_anchors
14
- dtype: int64
15
- - name: num_relationships_target
16
- dtype: int64
17
- - name: num_relationships_anchors
18
- dtype: int64
19
- - name: granularity
20
- dtype: string
21
- - name: ann_id
22
- dtype: string
23
- - name: object_ids
24
- list: int64
25
- - name: attribute_type_color
26
- dtype: bool
27
- - name: attribute_type_size
28
- dtype: bool
29
- - name: attribute_type_shape
30
- dtype: bool
31
- - name: attribute_type_number
32
- dtype: bool
33
- - name: attribute_type_material
34
- dtype: bool
35
- - name: attribute_type_texture
36
- dtype: bool
37
- - name: attribute_type_function
38
- dtype: bool
39
- - name: attribute_type_style
40
- dtype: bool
41
- - name: attribute_type_text_label
42
- dtype: bool
43
- - name: attribute_type_state
44
- dtype: bool
45
- - name: relationship_type_near
46
- dtype: bool
47
- - name: relationship_type_far
48
- dtype: bool
49
- - name: relationship_type_viewpoint_dependent
50
- dtype: bool
51
- - name: relationship_type_vertical
52
- dtype: bool
53
- - name: relationship_type_contain
54
- dtype: bool
55
- - name: relationship_type_arrangement
56
- dtype: bool
57
- - name: relationship_type_ordinal
58
- dtype: bool
59
- - name: relationship_type_comparison
60
- dtype: bool
61
- - name: coreferences
62
- dtype: bool
63
- - name: target_not_first_np
64
- dtype: bool
65
- - name: anchor_type_single
66
- dtype: bool
67
- - name: anchor_type_multiple
68
- dtype: bool
69
- - name: anchor_type_non_object
70
- dtype: bool
71
- - name: anchor_type_viewpoint
72
- dtype: bool
73
- - name: negation
74
- dtype: bool
75
- splits:
76
- - name: validation
77
- num_bytes: 72126
78
- num_examples: 349
79
- download_size: 45798
80
- dataset_size: 72126
81
- configs:
82
- - config_name: default
83
- data_files:
84
- - split: validation
85
- path: data/validation-*
86
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: mit
3
+ dataset_info:
4
+ features:
5
+ - name: scene_id
6
+ dtype: string
7
+ - name: ann_id
8
+ dtype: string
9
+ - name: object_name
10
+ dtype: string
11
+ - name: object_ids
12
+ list: int64
13
+ - name: description
14
+ dtype: string
15
+ - name: num_attributes_target
16
+ dtype: int64
17
+ - name: num_attributes_anchors
18
+ dtype: int64
19
+ - name: num_relationships_target
20
+ dtype: int64
21
+ - name: num_relationships_anchors
22
+ dtype: int64
23
+ - name: granularity
24
+ dtype: string
25
+ - name: attribute_type_color
26
+ dtype: bool
27
+ - name: attribute_type_size
28
+ dtype: bool
29
+ - name: attribute_type_shape
30
+ dtype: bool
31
+ - name: attribute_type_number
32
+ dtype: bool
33
+ - name: attribute_type_material
34
+ dtype: bool
35
+ - name: attribute_type_texture
36
+ dtype: bool
37
+ - name: attribute_type_function
38
+ dtype: bool
39
+ - name: attribute_type_style
40
+ dtype: bool
41
+ - name: attribute_type_text_label
42
+ dtype: bool
43
+ - name: attribute_type_state
44
+ dtype: bool
45
+ - name: relationship_type_near
46
+ dtype: bool
47
+ - name: relationship_type_far
48
+ dtype: bool
49
+ - name: relationship_type_viewpoint_dependent
50
+ dtype: bool
51
+ - name: relationship_type_vertical
52
+ dtype: bool
53
+ - name: relationship_type_contain
54
+ dtype: bool
55
+ - name: relationship_type_arrangement
56
+ dtype: bool
57
+ - name: relationship_type_ordinal
58
+ dtype: bool
59
+ - name: relationship_type_comparison
60
+ dtype: bool
61
+ - name: coreferences
62
+ dtype: bool
63
+ - name: target_not_first_np
64
+ dtype: bool
65
+ - name: anchor_type_single
66
+ dtype: bool
67
+ - name: anchor_type_multiple
68
+ dtype: bool
69
+ - name: anchor_type_non_object
70
+ dtype: bool
71
+ - name: anchor_type_viewpoint
72
+ dtype: bool
73
+ - name: negation
74
+ dtype: bool
75
+ splits:
76
+ - name: validation
77
+ num_bytes: 72126
78
+ num_examples: 349
79
+ download_size: 45798
80
+ dataset_size: 72126
81
+ configs:
82
+ - config_name: default
83
+ data_files:
84
+ - split: validation
85
+ path: data/validation-*
86
+ ---
87
+
88
+ # ViGiL3D Dataset
89
+ This repository contains the ViGiL3D benchmark dataset, introduced in [ViGiL3D: A Linguistically Diverse Dataset for 3D Visual Grounding](https://3dlg-hcvc.github.io/vigil3d/).
90
+
91
+ Prompts for both ScanNet and ScanNet++ are included in the same split.
92
+
93
+ ## Dataset Attributes
94
+
95
+ The key attributes are as follows:
96
+
97
+ **scene_id: (str)** the scene identifier as defined in [ScanNetv2](http://www.scan-net.org/) and [ScanNet++](https://kaldir.vc.in.tum.de/scannetpp/). \
98
+ **ann_id: (int)** the global annotation ID. \
99
+ **object_name: (str)** the name of the target object(s). \
100
+ **description: (str)** the language description of the target object(s) in the scene. \
101
+ **object_ids: (List[int])** the target object ID(s) as defined in [ScanNetv2](http://www.scan-net.org/).
102
+
103
+ In addition, we include the following metadata attributes characterizing the language patterns present in each description:
104
+
105
+ **num_attributes_target (int)** number of attributes describing the target object in the description \
106
+ **num_attributes_anchors (int)** number of attributes describing anchor objects in the description \
107
+ **num_relationships_target (int)** number of relationships describing the target object in the description \
108
+ **num_relationships_anchors (int)** number of relationships describing only anchor objects in the description \
109
+ **granularity (string)** the specificity of the noun used to refer to the target (`generic`, `categorical`, or `fine-grained`) \
110
+ **attribute_type_color (bool)** true if the color of an object is included in the description \
111
+ **attribute_type_size (bool)** true if the size of an object is included in the description \
112
+ **attribute_type_shape (bool)** true if the shape of an object is included in the description \
113
+ **attribute_type_number (bool)** true if the number of an object is included in the description \
114
+ **attribute_type_material (bool)** true if the material of an object is included in the description \
115
+ **attribute_type_texture (bool)** true if the texture of an object is included in the description \
116
+ **attribute_type_function (bool)** true if the function of an object is included in the description \
117
+ **attribute_type_style (bool)** true if the style of an object is included in the description \
118
+ **attribute_type_text_label (bool)** true if a text label on an object is included in the description \
119
+ **attribute_type_state (bool)** true if the state of an object (e.g. `open`, `closed`) is included in the description \
120
+ **relationship_type_near (bool)** true if an object is described as _near_ another in the description \
121
+ **relationship_type_far (bool)** true if an object is described as _far_ from another in the description \
122
+ **relationship_type_viewpoint_dependent (bool)** true if an object is described directionally relative to another (e.g. `left`, `right`) in the description \
123
+ **relationship_type_vertical (bool)** true if an object is described above, below, or on top of another in the description \
124
+ **relationship_type_contain (bool)** true if an object is described within or inside of another in the description \
125
+ **relationship_type_arrangement (bool)** true if an object is described as part of a larger arrangement (e.g. `row`, `stack`) in the description \
126
+ **relationship_type_ordinal (bool)** true if an object is described with an ordinal position (e.g. `first`, `third`) in the description \
127
+ **relationship_type_comparison (bool)** true if an object is compared to others with respect to an attribute (e.g. taller, widest) in the description \
128
+ **coreferences (bool)** true if an object is named using a coreference (e.g. `it`) in the description \
129
+ **target_not_first_np (bool)** true if the target object is not the first noun phrase in the description \
130
+ **anchor_type_single (bool)** true if one of the anchor objects in the description is a single object \
131
+ **anchor_type_multiple (bool)** true if one of the anchor objects in the description refers to multiple objects \
132
+ **anchor_type_non_object (bool)** true if one of the anchor objects in the description refers to a non-object (e.g. a region of the room) \
133
+ **anchor_type_viewpoint (bool)** true if one of the anchor objects in the description refers to an agent \
134
+ **negation (bool)** true if the description incldues negation \
135
+
136
+ ## Citation
137
+ ```bibtex
138
+ @inproceedings{wang2025vigil3d,
139
+ author={Wang, Austin T and Gong, ZeMing and Chang, Angel X},
140
+ title={ViGiL3D: A linguistically diverse dataset for 3d visual grounding},
141
+ booktitle={Proceedings of the Conference of the Association for Computational Linguistics (ACL)},
142
+ month={July},
143
+ year={2025},
144
+ }
145
+ ```