Datasets:

Modalities:
Image
Text
Formats:
parquet
Languages:
English
Libraries:
Datasets
Dask
License:
Llabres commited on
Commit
1ec086b
·
1 Parent(s): 6ad3683

Updated Readme

Browse files
Files changed (1) hide show
  1. README.md +24 -6
README.md CHANGED
@@ -42,31 +42,49 @@ tags:
42
  This is the dataset for the [ICDAR 2025 Competition on Comics Understanding in the Era of Foundational Models](https://rrc.cvc.uab.es/?ch=31&com=introduction)
43
 
44
  The dataset contains five subtask or skills:
45
- - Sequence Filling
46
- - Character Coherence
47
- - Visual Closure
48
- - Text Closure
49
- - Caption Relevance
50
 
51
- ## Skills:
52
  ### Sequence Filling
 
 
53
  ![Sequence Filling](figures/seq_filling.png)
54
  Given a sequence of comic panels, a missing panel, and a set of option panels, the task is to select the panel that best fits the sequence.
 
55
 
56
  ### Character Coherence, Visual Closure, Text Closure
 
 
57
  ![Character Coherence](figures/closure.png)
58
  These skills require understanding the context sequence to then pick the best panel to continue the story, focusing on the characters, the visual elements, and the text:
59
  - Character Coherence: Given a sequence of comic panels, pick the panel from the two options that best continues the story in a coherent with the characters. Both options are the same panel, but the text in the speech bubbles is has been swapped.
60
  - Visual Closure: Given a sequence of comic panels, pick the panel from the options that best continues the story in a coherent way with the visual elements.
61
  - Text Closure: Given a sequence of comic panels, pick the panel from the options that best continues the story in a coherent way with the text. All options are the same panel, but with text in the speech retrieved from different panels.
 
62
 
63
  ### Caption Relevance
 
 
64
  ![Caption Relevance](figures/caption_relevance.png)
65
  Given a caption from the previous panel, select the panel that best continues the story.
 
66
 
67
  ## Loading the Data
 
 
 
 
 
 
 
 
 
 
 
 
 
68
  _coming soon_
69
 
 
 
70
  ## Summit Results and Leaderboard
71
  The competition is hosted in the [Robust Reading Competition website](https://rrc.cvc.uab.es/?ch=31&com=introduction) and the leaderboard is available [here](https://rrc.cvc.uab.es/?ch=31&com=evaluation).
72
 
 
42
  This is the dataset for the [ICDAR 2025 Competition on Comics Understanding in the Era of Foundational Models](https://rrc.cvc.uab.es/?ch=31&com=introduction)
43
 
44
  The dataset contains five subtask or skills:
 
 
 
 
 
45
 
 
46
  ### Sequence Filling
47
+ <details>
48
+ <summary>Task Description</summary>
49
  ![Sequence Filling](figures/seq_filling.png)
50
  Given a sequence of comic panels, a missing panel, and a set of option panels, the task is to select the panel that best fits the sequence.
51
+ </details>
52
 
53
  ### Character Coherence, Visual Closure, Text Closure
54
+ <details>
55
+ <summary>Task Description</summary>
56
  ![Character Coherence](figures/closure.png)
57
  These skills require understanding the context sequence to then pick the best panel to continue the story, focusing on the characters, the visual elements, and the text:
58
  - Character Coherence: Given a sequence of comic panels, pick the panel from the two options that best continues the story in a coherent with the characters. Both options are the same panel, but the text in the speech bubbles is has been swapped.
59
  - Visual Closure: Given a sequence of comic panels, pick the panel from the options that best continues the story in a coherent way with the visual elements.
60
  - Text Closure: Given a sequence of comic panels, pick the panel from the options that best continues the story in a coherent way with the text. All options are the same panel, but with text in the speech retrieved from different panels.
61
+ </details>
62
 
63
  ### Caption Relevance
64
+ <details>
65
+ <summary>Task Description</summary>
66
  ![Caption Relevance](figures/caption_relevance.png)
67
  Given a caption from the previous panel, select the panel that best continues the story.
68
+ </details>
69
 
70
  ## Loading the Data
71
+
72
+ ```python
73
+ from datasets import load_dataset
74
+
75
+ skill = "seq_filling" # "seq_filling", "char_coherence", "visual_closure", "text_closure", "caption_relevance"
76
+ split = "val" # "test"
77
+ dataset = load_dataset("VLR-CVC/ComPAP", skill, split=split)
78
+ ```
79
+
80
+ <details>
81
+ <summary>Map to single images</summary>
82
+ If your model can only process single images, you can render each sample as a single image:
83
+
84
  _coming soon_
85
 
86
+ </details>
87
+
88
  ## Summit Results and Leaderboard
89
  The competition is hosted in the [Robust Reading Competition website](https://rrc.cvc.uab.es/?ch=31&com=introduction) and the leaderboard is available [here](https://rrc.cvc.uab.es/?ch=31&com=evaluation).
90