Less Is Better: Fooling Scene Text Recognition with Minimal Perturbations.

Yikun Xu, Pengwen Dai, Xiaochun Cao

Browse the full ICONIP paper archive.