Author name: Juncheng Hu, Ziteng Xue, Guotao Liang, Anran Qi, Buyu Li, Sheng Wang, Dong Xu, Qian Yu

arXiv:2604.10940v1 Announce Type: new
Abstract: We introduce AmodalSVG, a new framework for amodal image vectorization that produces semantically organized and geometrically complete SVG representations from natural images. Existing vectorization meth…