Semi-supervised Adversarial Learning to Generate Photorealistic Face Images of New Identities from 3D Morphable Model

We propose a novel end-to-end semi-supervised adversarial framework to generate photorealistic face images of new identities with wide ranges of expressions, poses, and illuminations conditioned by a 3D morphable model. Previous adversarial style-transfer methods either supervise their networks with large volume of paired data or use unpaired data with a highly under-constrained two-way generative framework in an unsupervised fashion... (read more)

PDF Abstract ECCV 2018 PDF ECCV 2018 Abstract

Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
Face Verification IJB-A VGG + GANFaces TAR @ FAR=0.01 53.507% # 16
TAR @ FAR=0.001 18.768 # 3
Face Verification Labeled Faces in the Wild VGG + GANFaces Accuracy 94.9% # 20

Methods used in the Paper


METHOD TYPE
Dropout
Regularization
Dense Connections
Feedforward Networks
ReLU
Activation Functions
Max Pooling
Pooling Operations
Softmax
Output Functions
Convolution
Convolutions
VGG
Convolutional Neural Networks