Computer Graphics and Vision

SketchyCOCO dataset

Introduction:

  1. Including 14k scene sketch images and their corresponding ground truth images.
  2. Including ground-truth label maps of the instance segmentation.
  3. Including 20k foreground sketch images and their corresponding ground-truth images.
  4. Including 27k background sketch images and their corresponding ground-truth images.
[Dataset] [Paper]

Scene sketch colorization dataset

Introduction:

  1. Including 38k groups of text-based instance segmentation data for scene sketch.
  2. Including 4k groups of text-based sketch object colorization data.
  3. Including 20k groups of text-based background colorization data for scene sketch.
[Dataset] [Paper]

SketchyScene dataset

Introduction:

  1. Including 7k scene sketch images and their corresponding reference cartoon images.
  2. Including ground-truth label maps of the semantic and instance segmentation.
  3. Including all the sketch objects used to compose the scene sketches and their layout order.
[Dataset] [Paper]

Back to top