From Noise to Latent: Generating Gaussian Latents for INR-Based Image Compression
This addresses the problem of high decoding complexity in image compression for applications requiring efficient transmission, though it appears incremental as it builds on existing INR and E2E methods.
The paper tackles the inferior performance of implicit neural representation (INR)-based image compression compared to end-to-end methods by proposing a novel paradigm that generates image-specific latents from Gaussian noise, eliminating the need to transmit latent codes and achieving competitive rate-distortion performance on Kodak and CLIC datasets.
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.