why different output of the same image in two different test?

[guoxiaolu]

I have used the pretrained model you offered(https://pan.baidu.com/s/1slMv4yp), and just modify the model name and image name of your test code. But in two different test, I have got different output results, for example, {good:0.6, bad:0.4} {good:0.4, bad:0.6}. It makes me confused and expects your answers~

[The victory-lab of Besti]

deploy.protxt consists of some initialize information , you could drop out these parts.
such as
weight_filler {
type: "xavier"
}
bias_filler {
type: "constant"
value: 0.2
}

[guoxiaolu]

Thank you very much, I will check it

[Chang Hsiungwei]

@BestiVictory I have same problem , and I try your method, but finally the score is always 0.5 whatever picture is.

[guoxiaolu]

Unfortunately, I didn't get the correct result, maybe other reasons. Finally, we reimplemented Adobe's paper, and get a good result.@FormatFish, you can make a try, it's a regression problem and Adobe used multi-task regression to solve it.