Question

Suppose we have an input image which is $\begin{pmatrix} 3 & 1 & 2 \\ 1 & 2 & 1 \\ 2 & 1 & 1 \end{pmatrix}$ and a convolution kernel which is $\begin{pmatrix} 0 & 1 & 0 \\ 1 & 0 & 1 \\ 0 & 1 & 1 \end{pmatrix}$ Assume we set stride as 1 and padding as 0 (no padding), determine the output of the convolution.

Name: suppose we have an input image which is beginpmatrix 3 1 2 1 2 1 2 1 1 endpmatrix and a convolution kernel which is beginpmatrix 0 1 0 1 0 1 0 1 1 endpmatrix assume we set stride as 1 and pa 40602
Uploaded: 2023-06-15T16:44:03-08:00
Duration: 1 min 17 s
Channel: Akash M
Description: suppose we have an input image which is beginpmatrix 3 1 2 1 2 1 2 1 1 endpmatrix and a convolution kernel which is beginpmatrix 0 1 0 1 0 1 0 1 1 endpmatrix assume we set stride as 1 and pa 40602

          Suppose we have an input image which is $\begin{pmatrix} 3 & 1 & 2 \\ 1 & 2 & 1 \\ 2 & 1 & 1 \end{pmatrix}$ and a convolution kernel which is $\begin{pmatrix} 0 & 1 & 0 \\ 1 & 0 & 1 \\ 0 & 1 & 1 \end{pmatrix}$
Assume we set stride as 1 and padding as 0 (no padding), determine the output of the convolution.

suppose we have an input image which is beginpmatrix 3 1 2 1 2 1 2 1 1 endpmatrix and a convolution kernel which is beginpmatrix 0 1 0 1 0 1 0 1 1 endpmatrix assume we set stride as 1 and pa 40602

Added by Natalia A.

Computer Science and Information Technology

Trishna Knowledge Systems 2018 Edition

Instant Answer

Solved by Expert Akash M

06/15/2023

Step 1

Step 1: The input image is a $3 \times 3$ matrix: $I = \begin{pmatrix} 3 & 1 & 2 \\ 1 & 2 & 1 \\ 2 & 1 & 1 \end{pmatrix}$ The convolution kernel is a $3 \times 3$ matrix: $K = \begin{pmatrix} 0 & 1 & 0 \\ 1 & 0 & 1 \\ 0 & 1 & 1 \end{pmatrix}$ The stride is 1 and Show more…

Show all steps

Thanks for your feedback!

Suppose we have an input image which is $\begin{pmatrix} 3 & 1 & 2 \\ 1 & 2 & 1 \\ 2 & 1 & 1 \end{pmatrix}$ and a convolution kernel which is $\begin{pmatrix} 0 & 1 & 0 \\ 1 & 0 & 1 \\ 0 & 1 & 1 \end{pmatrix}$ Assume we set stride as 1 and padding as 0 (no padding), determine the output of the convolution.