PIE: Pseudo-Invertible Encoder

10/31/2021
by   Jan Jetze Beitler, et al.
0

We consider the problem of information compression from high dimensional data. Where many studies consider the problem of compression by non-invertible transformations, we emphasize the importance of invertible compression. We introduce new class of likelihood-based autoencoders with pseudo bijective architecture, which we call Pseudo Invertible Encoders. We provide the theoretical explanation of their principles. We evaluate Gaussian Pseudo Invertible Encoder on MNIST, where our model outperforms WAE and VAE in sharpness of the generated images.

READ FULL TEXT

Please sign up or login with your details

Forgot password? Click here to reset