eduzhai > Applied Sciences > Engineering >

3D Topology Transformation with Generative Adversarial Networks

  • king
  • (0) Download
  • 20210505
  • Save

... pages left unread,continue reading

Document pages: 9 pages

Abstract: Generation and transformation of images and videos using artificialintelligence have flourished over the past few years. Yet, there are only a fewworks aiming to produce creative 3D shapes, such as sculptures. Here we show anovel 3D-to-3D topology transformation method using Generative AdversarialNetworks (GAN). We use a modified pix2pix GAN, which we call Vox2Vox, totransform the volumetric style of a 3D object while retaining the originalobject shape. In particular, we show how to transform 3D models into two newvolumetric topologies - the 3D Network and the Ghirigoro. We describe how touse our approach to construct customized 3D representations. We believe thatthe generated 3D shapes are novel and inspirational. Finally, we compare theresults between our approach and a baseline algorithm that directly convert the3D shapes, without using our GAN.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×