dc11ec32be
SAE : WARNING, RETRAIN IS REQUIRED ! fixed model sizes from previous update. avoided bug in ML framework(keras) that forces to train the model on random noise.
Colombo2019-09-19 11:13:56 +04:00
b6b92bded0
converter: now writes a filename of current frame config, SAE: removed multiscale decoder, because it's not effective
Colombo2019-09-13 08:59:00 +04:00
bef4e5d33c
Converter: added new color transfer modes: mkl, mkl-m, idt, idt-m
Colombo2019-09-07 22:01:01 +04:00
bc0b25cdda
interactive converter now saves the session. if input frames are changed (amount or filenames) then interactive converter automatically starts a new session. if model is more trained then all frames will be recomputed again with their saved configs.
Colombo2019-08-30 16:19:41 +04:00
9f58d160a0
Converter: fix output names of merged files, now its 100% same as input, ConvertAvatar: fix input image after fix landmarks face align, VideoEd: video_from_sequence now uses pipe input to input any filenames instead of %.5d. formatted
Colombo2019-08-27 22:05:22 +04:00
23854ac8bc
removing landmarks of lips which are used in face aligning, and leave only corners of mouth landmarks, result: less scale jittering aligning that will be feeded into AE and produce more stable face before: https://i.imgur.com/gJaW5Y4.gifv after: https://i.imgur.com/Vq7gvhY.gifviperov2019-08-26 19:36:41 +04:00
63d30c49ae
fixes and optimizations, converters: added new option sharpen_mode and sharpen_amount
iperov2019-08-25 19:18:51 +04:00
c39ed9d9c9
updated pdf manuals for AVATAR model. Avatar converter: added super resolution option. All converters: super resolution DCSCN network is now replaced by RankSRGAN
iperov2019-08-25 17:43:52 +04:00
19c66286da
fix gen_warp_params now accepts any square images
iperov2019-08-25 14:06:33 +04:00
6c66495bcf
default avatar type option is now HEAD. Head produces stable result.
iperov2019-08-25 08:16:54 +04:00
d129b5dd7f
sort data for CAInitializerMP
iperov2019-08-25 07:50:27 +04:00
407ce3b1ca
Added interactive converter. With interactive converter you can change any parameter of any frame and see the result in real time.
iperov2019-08-24 12:57:29 +04:00
b72d5a3f9a
fixed error "Failed to get convolution algorithm" on some systems fixed error "dll load failed" on some systems Expanded eyebrows line of face masks. It does not affect mask of FAN-x converter mode.
iperov2019-08-11 11:17:22 +04:00
93ee889597
Identity is not the same thing as equality in Python (#291)
cclauss2019-06-27 19:10:22 +02:00
8484060e01
Trainer: added option for all models Enable autobackup? (y/n ?:help skip:%s) : Autobackup model files with preview every hour for last 15 hours. Latest backup located in model/<>_autobackups/01
iperov2019-06-20 10:42:55 +04:00
2a8dd788dc
SAE: added option 'Pretrain the model?', Pretrain the model with large amount of various faces. This technique may help to train the fake with overly different face shapes and light conditions of src/dst data. Face will be look more like a morphed. To reduce the morph effect, some model files will be initialized but not be updated after pretrain: LIAE: inter_AB.h5 DF: both decoders.h5. The longer you pretrain the model the more morphed face will look. After that, save and run the training again.
iperov2019-05-01 19:55:27 +04:00
Colombo