PRNet_PyTorch icon indicating copy to clipboard operation
PRNet_PyTorch copied to clipboard

after training by all 300WLP datasets. when i evaluate the model. the NME is much lower than the paper. what is the problem?

Open ReverseSystem001 opened this issue 6 years ago • 6 comments

after training by all 300WLP datasets. when i evaluate the model. the NME is much lower than the paper. what is the problem?

ReverseSystem001 avatar Nov 29 '19 07:11 ReverseSystem001

Screenshot from 2019-11-29 15-06-43

ReverseSystem001 avatar Nov 29 '19 07:11 ReverseSystem001

when you are training the model. Do you use the AFLW2000 dataset(test dataset) ? hope your replay

ReverseSystem001 avatar Nov 29 '19 07:11 ReverseSystem001

Hello, ReverseSystem001, Where are you find the evaluate code? I am running the code currently, but did not find !

cvmaker93 avatar Mar 13 '20 06:03 cvmaker93

write it by myself. you can look at the 3ddfa evaluation code. 892379162 邮箱:[email protected] 签名由 网易邮箱大师 定制 On 03/13/2020 14:31, YUFENGZHANG wrote: Hello, ReverseSystem001, Where are you find the evaluate code? I am running the code currently, but did not find ! — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.

ReverseSystem001 avatar Mar 13 '20 11:03 ReverseSystem001

when I train by all 300WLP datasets, There is an error :TypeError: cannot unpack non-iterable NoneType object. Do you know how can i solve this error?

deepdarkfans avatar Oct 26 '20 02:10 deepdarkfans

when I train by all 300WLP datasets, There is an error :TypeError: cannot unpack non-iterable NoneType object. Do you know how can i solve this error?

have you solved this problems? and how? thx.

xianzhongma avatar Nov 11 '20 12:11 xianzhongma