FanNet: A mesh convolution operator for learning dense maps

Günes Sucu, Sinan Kalkan, Yusuf Sahillioglu

Abstract: Highlights•We introduce FanNet as a simple and novel mesh convolution operator.•FanNet is based on a neighborhood concept in which neighbors of a vertex are gathered as the spokes of a fan.•FanNet is applied to the problem of shape correspondence.•FanNet works well in both supervised and unsupervised settings.•FanNet provides on-par or better performance than the state-of-the-art.

External IDs:dblp:journals/cg/SucuKS25