I’m trying to use the discriminator of a GAN as an unsupervised classifier, to replicate the evaluation of DCGAN as per the DCGAN paper:
To evaluate the quality of the representations learned by DCGANs for supervised tasks,
we train on Imagenet-1k and then use the discriminator’s convolutional features from all layers,
maxpooling each layers representation to produce a 4 × 4 spatial grid. These features are then
flattened and concatenated to form a 28672 dimensional vector and a regularized linear L2-SVM
classifier is trained on top of them
EDIT:
Never mind! Pretty sure I should just follow this post about extracting features from each layer.