Skip to main content
. 2019 Jan 14;7:4900112. doi: 10.1109/JTEHM.2019.2892970

FIGURE 7.

FIGURE 7.

Fine-tuning configurations: green block indicates the layers for training (a) MANNE-S6: fine-tuning only the last layer before output, (b) MANNE-AS: fine-tuning last layers of all stages, (c) MANNE-AS-S2C3-#: fine-tuning last layers of all stages as well as the 3rd convolutional layer of stage 2 with # number of iterations.