1D and 3D Generalizations

Convolutions in 2D and 1D

2D Conv example

• 每筆資料的 input shape: $(14,14,\text{n_channels})$
• $\text{n_channels}=3$
• 每個 filter size: $(5,5,\text{n_channels})$
• output shape: $(10,10,\text{n_filters})$

1D Conv example

• 每筆資料的 input shape: $(14,\text{n_channels})$
• $\text{n_channels}=1$
• 每個 filter size: $(5,\text{n_channels})$
• output shape: $(10,\text{n_filters})$

3D data

• 斷層掃描
• 電影、影片

3D Conv example

• input shape: $(14,14,14,n_c)$
• $n_c=1$
• filter size: $(5,5,5,n_c)$
• output shape: $(10,10,10,\text{n_filters})$