Student: What is the purpose of using $1$x$1$ convolution in

CNN $?$

Professor: $1$x$1$ convolution reduces the size of feature maps

before sending them to layers with more expensive filters

$(3$x$3,5$x$5).$ A $(1$x$1)$ convolution would be useful if you

wanted to reduce the number of channels while preserving the

spatial dimensions.

Is the Professor$$s answer true or false?