I'd like to understand how to determine the parameter of conditional gaussian distribution. Following is the network architecture of VUNET which learns the conditional gaussian distribution $q(z|x, \hat y)$ of appearance of human($z$) upon given two conditional inputs skeleton of body pose and groundTruth image($x$) of 128x128 resolution image

vunet architecture

I don't understand how $2\times2\times256$ vector and $1\times1$ vector can characterize $q(z|x, \hat y)$. Any advice to understand this point clearly?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.