GoodDoggo
DF Vagrant
Posted as requested by @jmustard and @ztutt2.
Model and Faceset download: https://mega.nz/#!aihWES5R!fhySO8VEEIfK8U8Jpazl4IT5tAJ2Gb1ZIAgN2yrH9-8
I included an additional aligned set the model has yet to train with and can be added to the main aligned folder if you decide to. As always make sure to scan through the images and remove any undesirable faces or false positives in case I missed any. I aimed to cover a wide range of expressions and angles rather than go for quantity alone. The second aligned set has more of these expressions and mainly come from a pizza eating video but may be undesirable according to your personal taste.
Model details:
========= Model Summary ==========
== ==
== Model name: H128 ==
== ==
== Current iteration: 487556 ==
== ==
==------- Model Options --------==
== ==
== batch_size: 28 ==
== lighter_ae: False ==
== sort_by_yaw: False ==
== random_flip: False ==
== pixel_loss: False ==
== ==
==--------- Running On ---------==
== ==
== Device index: 0 ==
== Name: Tesla K80 ==
== VRAM: 11.00GB ==
== ==
==================================
Model and Faceset download: https://mega.nz/#!aihWES5R!fhySO8VEEIfK8U8Jpazl4IT5tAJ2Gb1ZIAgN2yrH9-8
I included an additional aligned set the model has yet to train with and can be added to the main aligned folder if you decide to. As always make sure to scan through the images and remove any undesirable faces or false positives in case I missed any. I aimed to cover a wide range of expressions and angles rather than go for quantity alone. The second aligned set has more of these expressions and mainly come from a pizza eating video but may be undesirable according to your personal taste.
Model details:
========= Model Summary ==========
== ==
== Model name: H128 ==
== ==
== Current iteration: 487556 ==
== ==
==------- Model Options --------==
== ==
== batch_size: 28 ==
== lighter_ae: False ==
== sort_by_yaw: False ==
== random_flip: False ==
== pixel_loss: False ==
== ==
==--------- Running On ---------==
== ==
== Device index: 0 ==
== Name: Tesla K80 ==
== VRAM: 11.00GB ==
== ==
==================================