search

soubhiksanyal / RingNet

Learning to Regress 3D Face Shape and Expression from an Image without 3D Supervision
https://ringnet.is.tue.mpg.de
MIT License
821 stars 170 forks source link

Pose parameters for neck #25

Closed bing-jian closed 4 years ago

bing-jian commented 4 years ago

"RingNet outputs a subset of the FLAME parameters. These parameters are the first 100 shape parameters, first 50 expression parameters, and the pose parameters for the global rotation (FLAME pose parameters 1-3) and the jaw rotation (FLAME pose parameters 7-9). "

Originally posted by @TimoBolkart in https://github.com/soubhiksanyal/RingNet/issues/7#issuecomment-503638709

Wondering what are the FLAME pose parameters 4-6. Are these for the neck which is also a joint in the FLAME model? Does RingNet outputs pose parameters for neck?

Thanks, Bing

bing-jian commented 4 years ago

Found the answer in the Ringnet paper Section 3.5 Implementation details: The neck and eyeball rotations of FLAME are not regressed since the facial landmarks do not impose any constraints on the neck.

The regression network consists of two fully-connected layers of dimension 512 with ReLu activation and dropout, followed by a final linear fully-connected layer with 159-dimensional output. To this 159-dimensional output vector we concatenate the camera, pose, shape, and expression parameters. The first three elements represent scale and 2D image translation.

The following 6 elements are the global rotation and jaw rotation, each in axis-angle representation. The neck and eyeball rotations of FLAME are not regressed since the facial landmarks do not impose any constraints on the neck. The next 100 elements are the shape parameters, followed by 50 expression parameters of FLAME.