jingqun commited on
Commit
bfb0b0b
·
verified ·
1 Parent(s): 9480872

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -2
README.md CHANGED
@@ -24,7 +24,7 @@ size_categories:
24
  The dataset is oriented toward visual question answering of multilingual text scenes in nine languages, including Korean, Japanese, Italian, Russian, Deutsch, French, Thai, Arabic, and Vietnamese. The question-answer pairs are labeled by native annotators following a series of rules. A comprehensive description of the dataset can be found in the paper [MTVQA](https://arxiv.org/pdf/2405.11985).
25
 
26
  ## - Image Distribution
27
- <table style="width:50%;">
28
  <tr>
29
  <td><b>KO</b></td>
30
  <td><b>JA</b></td>
@@ -50,7 +50,7 @@ The dataset is oriented toward visual question answering of multilingual text sc
50
  <td>1139</td>
51
  </tr>
52
  <tr>
53
- <td><b>Test images</b></td>
54
  <td>250</td>
55
  <td>250</td>
56
  <td>250</td>
 
24
  The dataset is oriented toward visual question answering of multilingual text scenes in nine languages, including Korean, Japanese, Italian, Russian, Deutsch, French, Thai, Arabic, and Vietnamese. The question-answer pairs are labeled by native annotators following a series of rules. A comprehensive description of the dataset can be found in the paper [MTVQA](https://arxiv.org/pdf/2405.11985).
25
 
26
  ## - Image Distribution
27
+ <table style="width:60%;">
28
  <tr>
29
  <td><b>KO</b></td>
30
  <td><b>JA</b></td>
 
50
  <td>1139</td>
51
  </tr>
52
  <tr>
53
+ <td><b>Test Images</b></td>
54
  <td>250</td>
55
  <td>250</td>
56
  <td>250</td>