Abstract: Zero-shot learning for visual recognition, which approaches identifying unseen categories through a shared visual-semantic function learned on the seen categories and is expected to well ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results