How does the random flip works in Losses for pseudo exemplar pairs

[neew123]

I try to train CoCosNet in my own dataset,but i find it is random to get the right result because of the probability of flip?.So I want to ask How does the random flip works in Losses for pseudo exemplar pairs?if I choose the exempalr no filp, this difference can influence the final result? Thanks for your answering.

[panzhang0212]

Random flip is a kind of method to generate pseudo exemplar pairs. Other augmentations, e.g. geometry affine translation, random flow, could also be used to generate pseudo exemplar pairs. During the test phase, flip or not could not influence the final result.

[neew123]

Random flip is a kind of method to generate pseudo exemplar pairs. Other augmentations, e.g. geometry affine translation, random flow, could also be used to generate pseudo exemplar pairs. During the test phase, flip or not could not influence the final result.

Thanks for your answering,actually ,when I use CoCosNet train on my own edge-to-image dataset,The first twenty iterations it can generate good results,but when after twenty iterations,the result is the same as exemplar but not use edge image.So,I want to ask If the probablity of flip can affect the final results.