Overcoming the limits of rare disease matching using facial phenotypic descriptors

Access to source code and data

Please visit the download page in GestaltMatcher Database for accessing the code and data for training.

GestaltMatcher Architecture

Gestaltmatcher architecture

Name Type Filter size / Stride Output size
Conv 11Convolution3x3/1100x100x32
Conv 12Convolution3x3/1100x100x64
Pool1Max pooling2x2/250x50x64
Conv 21Convolution3x3/150x50x64
Conv 22Convolution3x3/150x50x128
Pool2Max pooling2x2/225x25x128
Conv 31Convolution3x3/125x25x96
Conv 32Convolution3x3/125x25x192
Pool3Max pooling2x2/213x13x192
Conv 41Convolution3x3/113x13x128
Conv 42Convolution3x3/113x13x256
Pool4Max pooling2x2/27x7x256
Conv 51Convolution3x3/17x7x160
Conv 52Convolution3x3/17x7x320
Pool5Avg pooling7x7/11x1x320
DropoutDroupout (50%)-1x1x320
FC6Fully connected-# Classes
CostSoftmax-# Classes
* Important note #1: Every convolutional layer is followed by a batch normalization and a Relu layer
* Important note #2: This architecture is almost identical to CASIA-Webface face recognition architecture (see table from original paper . )