Few-Shot Class-Incremental Learning (FSCIL) is a novel problem setting for incremental learning, where a unified classifier is incrementally learned for new classes with very few training samples. In ...
(Note: the SphereFace implementation is not exactly as described in their paper but instead uses the 'trick' presented in the ArcFace paper to use arccosine instead of the double angle formula) There ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results