sal4ahm commited on
Commit
fc0f3c1
1 Parent(s): 8596689

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +16 -18
README.md CHANGED
@@ -25,6 +25,7 @@ The dataset is organized into the following folders:
25
  - `jsons3`: Contains the JSON files corresponding to the images in the `images3` folder.
26
 
27
  - **QA Files**
 
28
  - `qa`: Contains the QA files corresponding to the images in the `images` folder.
29
  - `qa2`: Contains the QA files corresponding to the images in the `images2` folder.
30
  - `qa3`: Contains the QA files corresponding to the images in the `images3` folder.
@@ -32,11 +33,12 @@ The dataset is organized into the following folders:
32
  ### File Details
33
 
34
  - **Images**: JPEG files named in the format `PMCxxxxxx_abc.jpg`, where `xxxxxx` represents the PubMed Central ID and `abc` represents an identifier specific to the image.
35
- - **JSON Files**: JSON files named in the same format as the images. Each JSON file is a list of question blocks associated with the corresponding image.
 
36
 
37
- #### JSON Structure
38
 
39
- Each JSON file contains a list of question blocks in the following format:
40
 
41
  ```json
42
  [
@@ -61,8 +63,6 @@ Each JSON file contains a list of question blocks in the following format:
61
  ]
62
  ```
63
 
64
- - **QA Files**: Contain additional question-answer metadata relevant to the dataset.
65
-
66
  ### Dataset Loader
67
 
68
  To facilitate loading and using the dataset, we provide a custom dataset loader script, `dataset.py`. This script defines a PyTorch `Dataset` class to handle loading, preprocessing, and batching of the images and question-answer pairs.
@@ -85,22 +85,20 @@ To facilitate loading and using the dataset, we provide a custom dataset loader
85
  from dataset import RQADataset
86
  from torch.utils.data import DataLoader
87
 
88
- # Define the data configuration
89
- class DataConfig:
90
- img_dir = '/home/jupyter/data/RQA/images' # Update with actual image directory path
91
- json_dir = '/home/jupyter/data/RQA/jsons' # Update with actual JSON directory path
92
- filter_list = '/home/jupyter/data/RQA_V0/test_filenames.txt' # Path to the file containing test filenames
93
- train = False # Set to True for training, False for testing
94
 
95
- # Initialize dataset
96
- dataset = RQADataset(DataConfig)
 
 
97
 
98
- # Initialize DataLoader
99
- dataloader = DataLoader(dataset, batch_size=4, collate_fn=RQADataset.custom_collate)
100
 
101
- # Iterate through the dataset
102
- for batch in dataloader:
103
- print(batch)
 
104
  ```
105
 
106
  ### Citation
 
25
  - `jsons3`: Contains the JSON files corresponding to the images in the `images3` folder.
26
 
27
  - **QA Files**
28
+ These are the QA created in our proposed dataset.
29
  - `qa`: Contains the QA files corresponding to the images in the `images` folder.
30
  - `qa2`: Contains the QA files corresponding to the images in the `images2` folder.
31
  - `qa3`: Contains the QA files corresponding to the images in the `images3` folder.
 
33
  ### File Details
34
 
35
  - **Images**: JPEG files named in the format `PMCxxxxxx_abc.jpg`, where `xxxxxx` represents the PubMed Central ID and `abc` represents an identifier specific to the image.
36
+ - **JSON Files**: JSON files named in the same format as the images. These are groundtruth annotations from the https://chartinfo.github.io challenge, they provide annotations for chart type, text(OCR), text location, text type (axis/tick/legend), data used to plot the chart.
37
+ - **QA Files**: QA files named in the same format as the images. Each QA file is a list of question blocks associated with the corresponding image we created in our proposed dataset.
38
 
39
+ #### QA Structure
40
 
41
+ Each QA file contains a list of question blocks in the following format:
42
 
43
  ```json
44
  [
 
63
  ]
64
  ```
65
 
 
 
66
  ### Dataset Loader
67
 
68
  To facilitate loading and using the dataset, we provide a custom dataset loader script, `dataset.py`. This script defines a PyTorch `Dataset` class to handle loading, preprocessing, and batching of the images and question-answer pairs.
 
85
  from dataset import RQADataset
86
  from torch.utils.data import DataLoader
87
 
88
+ dataset = RQADataset(data_dir='.', split='train') # split='test' for RQA9357 split used in the paper
 
 
 
 
 
89
 
90
+ # Test loading a single item
91
+ print(f"Number of samples in dataset: {len(dataset)}")
92
+ sample = dataset[0]
93
+ print("Sample data:", sample)
94
 
95
+ # Initialize DataLoader
96
+ dataloader = DataLoader(dataset, batch_size=4, collate_fn=RQADataset.custom_collate)
97
 
98
+ # Test DataLoader
99
+ for batch in dataloader:
100
+ print("Batch data:", batch)
101
+ break # Load only one batch for testing
102
  ```
103
 
104
  ### Citation