3D Shape Reconstruction from Free-Hand Sketches

  • 2020-06-17 07:43:10
  • Jiayun Wang, Jierui Lin, Qian Yu, Runtao Liu, Yubei Chen, Stella X. Yu
  • 28


Sketches are the most abstract 2D representations of real-world objects.Although a sketch usually has geometrical distortion and lacks visual cues,humans can effortlessly envision a 3D object from it. This indicates thatsketches encode the appropriate information to recover 3D shapes. Althoughgreat progress has been achieved in 3D reconstruction from distortion-free linedrawings, such as CAD and edge maps, little effort has been made to reconstruct3D shapes from free-hand sketches. We pioneer to study this task and aim toenhance the power of sketches in 3D-related applications such as interactivedesign and VR/AR games. Further, we propose an end-to-end sketch-based 3Dreconstruction framework. Instead of well-used edge maps, synthesized sketchesare adopted as training data. Additionally, we propose a sketch standardizationmodule to handle different sketch styles and distortions. With extensiveexperiments, we demonstrate the effectiveness of our model and its stronggeneralizability to various free-hand sketches.


Quick Read (beta)

loading the full paper ...