roopalgarg commited on
Commit
337b60f
1 Parent(s): 690b6c2

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +6 -5
README.md CHANGED
@@ -64,9 +64,9 @@ We validate the framework through evaluations focused on the quality of the data
64
 
65
  This Data Card describes a mixture of human annotated and machine generated data intended to help create and capture rich, hyper-detailed image descriptions.
66
 
67
- IIW dataset has two parts: human annotations and model outputs. The main purposes of this dataset are:
68
- (1) to provide samples from SoTA human authored outputs to promote discussion on annotation guidelines to further improve the quality
69
- (2) to provide human SxS results and model outputs to promote development of automatic metrics to mimic human SxS judgements.
70
 
71
  ### Supported Tasks
72
 
@@ -82,6 +82,8 @@ English
82
 
83
  ### Data Fields
84
 
 
 
85
  IIW:
86
  - `image/key`
87
  - `image/url`
@@ -98,7 +100,7 @@ DCI:
98
  - `image`
99
  - `image/url`
100
  - `ex_id`
101
- - `IIW`: Human generated image description
102
  - `metrics/Comprehensiveness`
103
  - `metrics/Specificity`
104
  - `metrics/Hallucination`
@@ -107,7 +109,6 @@ DCI:
107
 
108
  DOCCI:
109
  - `image`
110
- - `image/url`
111
  - `image/thumbnail_url`
112
  - `IIW`: Human generated image description
113
  - `DOCCI`: Image description from DOCCI
 
64
 
65
  This Data Card describes a mixture of human annotated and machine generated data intended to help create and capture rich, hyper-detailed image descriptions.
66
 
67
+ IIW dataset has two parts: human annotations and model outputs. The main purposes of this dataset are:
68
+ 1) to provide samples from SoTA human authored outputs to promote discussion on annotation guidelines to further improve the quality
69
+ 2) to provide human SxS results and model outputs to promote development of automatic metrics to mimic human SxS judgements.
70
 
71
  ### Supported Tasks
72
 
 
82
 
83
  ### Data Fields
84
 
85
+ For details on the datasets and output keys, please refer to our [GitHub data](https://github.com/google/imageinwords/tree/main/datasets) page inside the individual folders.
86
+
87
  IIW:
88
  - `image/key`
89
  - `image/url`
 
100
  - `image`
101
  - `image/url`
102
  - `ex_id`
103
+ - `IIW`: Human authored image description
104
  - `metrics/Comprehensiveness`
105
  - `metrics/Specificity`
106
  - `metrics/Hallucination`
 
109
 
110
  DOCCI:
111
  - `image`
 
112
  - `image/thumbnail_url`
113
  - `IIW`: Human generated image description
114
  - `DOCCI`: Image description from DOCCI