Many recent papers labeled as “reconstruction” are effectively performing n-way decoding: they decode a category and then use a generative model to produce a visually plausible sample inside the predicted class. This can look convincing, but remains a restricted classification problem. Similar n-way decoding setups have long been standard in the MVPA literature and can achieve quite high performance when the candidate set is limited.
«Модная Заметка. Писал вам недавно про растущую популярность каракуля. И вот получайте: видим вещи из каракуля на подиуме у лучших из лучших! А мы еще до Парижа "не доехали"!» — написал он.
，详情可参考PDF资料
Will tech trump tradition at bakers and biscuit makers?
France 24 live updates