Skip to main content
[Preprint]. 2023 Aug 26:2023.08.04.551906. Originally published 2023 Aug 9. [Version 2] doi: 10.1101/2023.08.04.551906

Fig. 2. The training results of CAE-2 module.

Fig. 2.

As illustrated by the generated images for one real sample image with 10, 40, 70 and 100 epochs respectively. The trained model is prone to converge at epoch 100.