**StyleMapGAN** is a generative adversarial network for real-time image editing. The intermediate latent space has spatial dimensions, and a spatially variant modulation replaces [AdaIN](https://paperswithcode.com/method/adaptive-instance-normalization). It aims to make the embedding through an encoder more accurate than existing optimization-based methods while maintaining the properties of GANs.

**Discriminative Regularization** is a regularization technique for [variational autoencoders](https://paperswithcode.com/methods/category/likelihood-based-generative-models) that uses representations from discriminative classifiers to augment the [VAE](https://paperswithcode.com/method/vae) objective function (the lower bound) corresponding to a generative model. Specifically, it encourages the model’s reconstructions to be close to the data example in a representation space defined by the hidden layers of highly-discriminative, neural network based classifiers.

Discriminative Regularization

Discriminative Regularization for Generative Models

StyleMapGAN

Exploiting Spatial Dimensions of Latent in GAN for Real-time Image Editing

**DV3 Convolution Block** is a convolutional block used for the [Deep Voice 3](https://paperswithcode.com/method/deep-voice-3) text-to-speech architecture. It consists of a 1-D [convolution](https://paperswithcode.com/method/convolution) with a gated linear unit and a [residual connection](https://paperswithcode.com/method/residual-connection). In the Figure, $c$ denotes the dimensionality of the input. The convolution output of size $2 \cdot c$ is split into equal-sized portions: the gate vector and the input vector. A scaling factor $\sqrt{0.5}$ is used to ensure that we preserve the input variance early in training. The gated linear unit provides a linear path for the gradient flow, which alleviates the vanishing gradient issue for stacked convolution blocks while retaining non-linearity. To introduce speaker-dependent control, a speaker-dependent embedding is added as a bias to the convolution filter output, after a softsign function. The authors use the softsign nonlinearity because it limits the range of the output while also avoiding the saturation problem that exponential based nonlinearities sometimes exhibit. Convolution filter weights are initialized with zero-mean and unit-variance activations throughout the entire network.

Source	Exploiting Spatial Dimensions of Latent in GAN for Real-time Image Editing
Year	2000
Data Source	CC BY-SA - https://paperswithcode.com