OpenSketch

OpenSketch is a SketchDataset loader for the OpenSketch dataset, which contains more than 400 product design sketches representing 12 man-made objects drawn by 7 to 15 product designers of varying expertise. The dataset is designed for sketch-based modeling, nonphotorealistic rendering, and sketch filtering tasks. Each sketch is stored as an SVG file containing vector paths.

Source: datasets/opensketch.py

Data Format

Each sample is an SVG file containing:

  • <path> elements with d attributes defining drawing paths

  • viewBox attribute defining canvas dimensions

The loader parses SVG paths and converts them into cubic Bézier curves. Supported SVG path commands include:

  • Move (M): Starting point

  • Cubicezier (C): Cubic Bézier curves

  • Line (L): Straight lines (converted to cubic Bézier curves)

Each path is converted into a Path composed of cubic Bézier Curves. The canvas dimensions are extracted from the SVG viewBox attribute.

Directory Layout

After download and extraction, the dataset is expected under:

<root>/
  open-sketch_svgs/
    *.svg
  sketches_svg.zip
  .metadata.parquet

  • All sketch files are stored as .svg files in the opensketch_svgs/ directory.

  • The original archive sketches_svg.zip is kept in the root directory for integrity checking.

Code

from sketchkit.datasets import OpenSketch

ds = OpenSketch(cislab_source=True)

sketch = ds[0]
print(sketch.width, sketch.height)
print(sketch.path_num, sketch.curve_num)

Arguments

  • cislab_source: Selects the download source. If True, the dataset is downloaded from the CISLAB CDN mirror; otherwise it is downloaded from official host.

    • CISLAB mirror: https://cislab.hkust-gz.edu.cn/projects/sketchkit/datasets/OpenSketch/sketches_svg.zip

    • Official host: https://repo-sam.inria.fr/d3/OpenSketch/Data/sketches/sketches_svgs.zip