Wavelet-based Variational Autoencoders for High-Resolution Image Generation
This addresses the issue of generating high-resolution images for applications in computer vision and generative modeling, though it appears incremental as it builds on existing VAE frameworks.
The paper tackled the problem of blurry image generation in conventional Variational Autoencoders by introducing a wavelet-based approach (Wavelet-VAE) that constructs the latent space using multi-scale Haar wavelet coefficients, resulting in improved visual fidelity and recovery of higher-resolution details on datasets like CIFAR-10.
Variational Autoencoders (VAEs) are powerful generative models capable of learning compact latent representations. However, conventional VAEs often generate relatively blurry images due to their assumption of an isotropic Gaussian latent space and constraints in capturing high-frequency details. In this paper, we explore a novel wavelet-based approach (Wavelet-VAE) in which the latent space is constructed using multi-scale Haar wavelet coefficients. We propose a comprehensive method to encode the image features into multi-scale detail and approximation coefficients and introduce a learnable noise parameter to maintain stochasticity. We thoroughly discuss how to reformulate the reparameterization trick, address the KL divergence term, and integrate wavelet sparsity principles into the training objective. Our experimental evaluation on CIFAR-10 and other high-resolution datasets demonstrates that the Wavelet-VAE improves visual fidelity and recovers higher-resolution details compared to conventional VAEs. We conclude with a discussion of advantages, potential limitations, and future research directions for wavelet-based generative modeling.