Gaussian smoothin row vector 1 2 ] approximates a gaussian kernel with standard deviation sqrt(1/2). (note, we are ignoring the normalization constant to keep all values integers). in the following, report the results of doing "full" convolution with zero-padding, that is, a value is computed for every partial overlap of the two kernels (the result is then bigger than either input kernel). 2a (3pts) what is [1 2 1] convolved with [1 2 1] 2b (3pts) what is [ 1 2 i ] convolved with | 1 2c (2pt) what would the standard deviation be for the gaussian computed in part 2a (if the values were normalized properly to add up to) enter values rounded to 2 decimal places sigma- 2d (2pt) what would the standard deviation(s) be for the gaussian computed in part 2b (if the values were normalized properly to add up to 1)? enter values rounded to 2 decimal places sigmax sigmay