From Noise to Latent: Generating Gaussian Latents for INR-Based Image Compression

TL;DR

This paper introduces a novel image compression method that generates image-specific latents directly from Gaussian noise, eliminating the need to transmit latent codes while maintaining competitive quality.

Contribution

It proposes a new paradigm of generating latents from Gaussian noise using a shared seed and a parameter prediction module, pioneering this approach in learned image compression.

Findings

Achieves competitive rate-distortion performance on Kodak and CLIC datasets.

Eliminates the need to transmit latent codes, reducing decoding complexity.

First work to explore Gaussian latent generation for learned image compression.

Abstract

Recent implicit neural representation (INR)-based image compression methods have shown competitive performance by overfitting image-specific latent codes. However, they remain inferior to end-to-end (E2E) compression approaches due to the absence of expressive latent representations. On the other hand, E2E methods rely on transmitting latent codes and requiring complex entropy models, leading to increased decoding complexity. Inspired by the normalization strategy in E2E codecs where latents are transformed into Gaussian noise to demonstrate the removal of spatial redundancy, we explore the inverse direction: generating latents directly from Gaussian noise. In this paper, we propose a novel image compression paradigm that reconstructs image-specific latents from a multi-scale Gaussian noise tensor, deterministically generated using a shared random seed. A Gaussian Parameter Prediction (GPP) module estimates the distribution parameters, enabling one-shot latent generation via reparameterization trick. The predicted latent is then passed through a synthesis network to reconstruct the image. Our method eliminates the need to transmit latent codes while preserving latent-based benefits, achieving competitive rate-distortion performance on Kodak and CLIC dataset. To the best of our knowledge, this is the first work to explore Gaussian latent generation for learned image compression.