抄録
CH-014
新出に対応する深層学習を用いたメタ認知に基づく画像認識
新出に対応する深層学習を用いたメタ認知に基づく画像認識
Open Set Recognition is the task of classifying test data that differ in some respect from the data that are available during training.
Identifying samples from currently unknown classes is hence an essential step in visual object recognition.
In this paper, I propose a new approach for open set recognition, Deep Meta-Recognition Networks (DMRN).
I evaluate the resulting DMRN considering the difference between the category of known class and the category of unknown class.
Identifying samples from currently unknown classes is hence an essential step in visual object recognition.
In this paper, I propose a new approach for open set recognition, Deep Meta-Recognition Networks (DMRN).
I evaluate the resulting DMRN considering the difference between the category of known class and the category of unknown class.
