Hugging Face's logo

Datasets:
13point5
/
lineex-test

Tasks:
Image Classification
Object Detection
Image-to-Text
Modalities:
Image
Text
Formats:
parquet
Size:
10K - 100K
Libraries:
Datasets
Dask
Polars
Dataset card Data Studio Files
xet
 Community
lineex-test / README.md
13point5's picture
13point5
Add dataset card
6ca0bc6 verified
preview code
|
raw
history blame contribute delete
1.79 kB
metadata 
pretty_name: LineEX Test Split
task_categories:
  - image-classification
  - object-detection
  - image-to-text
size_categories:
  - 10K<n<100K
configs:
  - config_name: default
    data_files:
      - split: test
        path: data/test-*

LineEX Test Split

This dataset is a structured Hugging Face mirror of the test split released with the LineEX paper, "LineEX: Data Extraction from Scientific Line Charts" (WACV 2023).

  • Repo: 13point5/lineex-test
  • Split: test
  • Rows: 20,000
  • Source format: PNG images plus COCO-style chart-element and line annotations

Columns

  • image_id
  • file_name
  • image
  • width
  • height
  • data_type
  • chart_elements
  • lines

chart_elements fields

  • annotation_id
  • category_id
  • category_name
  • bbox_xywh
  • area
  • text
  • line_id

lines fields

  • annotation_id
  • category_id
  • category_name
  • line_name
  • polyline_xy
  • raw_series_xy
  • area

Chart element category mapping

  • 0: Legend
  • 1: ValueAxisTitle
  • 2: ChartTitle
  • 3: CategoryAxisTitle
  • 4: PlotArea
  • 5: InnerPlotArea
  • 6: XY_Ticks
  • 7: LegendMarker
  • 8: LegendText
  • 9: LegendElement

Notes

  • The original cls_anno.json category list is incomplete relative to the annotation IDs. This mirror uses the 10-class mapping defined in the LineEX codebase.
  • line_id links legend marker annotations to line annotations within the same image.
  • lines.line_name is derived from the legend structure and may be null when no legend-text association can be inferred for a series.
  • This repo mirrors only the test split.

License

This mirror does not assert a new license for the data. Verify redistribution terms with the original dataset authors before making the dataset public.