[4I2-J-2] Machine learning: uncertainty of targets

[4I2-J-2-02] Deep learning based on GAIN for mimesis

〇Kai Tomono¹, Hajime Nobuhara¹, Shin Kawai¹ (1. University of Tsukuba)

Keywords:Anolis carolinensis, bias

The bias of the training image data is considered to have a great in uence on the identi cation of mimicable
organisms. The purpose of this study is to examine the effectiveness of Guided Attention Inference Networks
(GAIN) for recognizing the Anolis carolinensis. We also con rmed the improvement of classi cation accuracy from
the veri cation results.