IntroVAE: Introspective Variational Autoencoders for Photographic Image Synthesis
This work addresses the challenge of stable and efficient high-quality image generation for applications in computer vision and graphics, representing a novel hybrid approach rather than an incremental improvement.
The authors tackled the problem of synthesizing high-resolution photographic images by introducing IntroVAE, a model that integrates VAE and GAN frameworks in a single-stream architecture, achieving photo-realistic results comparable to state-of-the-art GANs on datasets like CELEBA at 1024^2 resolution.
We present a novel introspective variational autoencoder (IntroVAE) model for synthesizing high-resolution photographic images. IntroVAE is capable of self-evaluating the quality of its generated samples and improving itself accordingly. Its inference and generator models are jointly trained in an introspective way. On one hand, the generator is required to reconstruct the input images from the noisy outputs of the inference model as normal VAEs. On the other hand, the inference model is encouraged to classify between the generated and real samples while the generator tries to fool it as GANs. These two famous generative frameworks are integrated in a simple yet efficient single-stream architecture that can be trained in a single stage. IntroVAE preserves the advantages of VAEs, such as stable training and nice latent manifold. Unlike most other hybrid models of VAEs and GANs, IntroVAE requires no extra discriminators, because the inference model itself serves as a discriminator to distinguish between the generated and real samples. Experiments demonstrate that our method produces high-resolution photo-realistic images (e.g., CELEBA images at \(1024^{2}\)), which are comparable to or better than the state-of-the-art GANs.