Abstract: Methods, apparatus, systems achieve disentanglement in semantic editing using GANs-based models. Self-corrected (low-density) latent code samples are projected in the original latent space and the editing directions corrected though relearning based on resulting high-density and low-density regions in the amended latent space. Leveraging the original meaningful directions and semantic region-specific layers, operations interpolate the original latent codes to generate images with minority combinations of attributes, then inverts these samples back to the original latent space. In accordance with embodiments, the operations can apply to preexisting methods that learn meaningful latent directions. Attribute disentanglement is improved with small amounts of low-density region samples added.

Abstract: Methods, apparatus and techniques herein relates to determining directions in GAN latent space and obtaining disentangled controls over GAN output semantics, for example, to enable use of such to generating synthesized images such as for use to train another model or create an augmented reality The methods, apparatus and techniques herein, in accordance with embodiments, utilize the gradient directions of auxiliary networks to control semantics in GAN latent codes. It is shown that minimal amounts of labelled data with sizes as small as 60 samples can be used, which data can be obtained quickly with human supervision. It is also shown herein, in accordance with embodiments, to select important latent code channels with masks during manipulation, resulting in more disentangled controls.