GaINeR: Geometry-Aware Implicit Network Representation

TL;DR

GaINeR introduces a geometry-aware implicit neural network for 2D image representation that enables high-quality reconstruction, flexible editing, and integration with physical simulations by combining Gaussian distributions with neural networks.

Contribution

This work presents GaINeR, a novel framework that incorporates trainable Gaussian distributions into INRs for interpretable geometric structure and enhanced image editing capabilities.

Findings

Achieves state-of-the-art reconstruction quality.

Supports geometry-consistent transformations and super-resolution.

Enables depth-guided editing through 3D representation.

Abstract

Implicit Neural Representations (INRs) are widely used for modeling continuous 2D images, enabling high-fidelity reconstruction, super-resolution, and compression. Architectures such as SIREN, WIRE, and FINER demonstrate their ability to capture fine image details. However, conventional INRs lack explicit geometric structure, limiting local editing, and integration with physical simulation. To address these limitations, we propose GaINeR (Geometry-Aware Implicit Neural Representation for Image Editing), a novel framework for 2D images that combines trainable Gaussian distributions with a neural network-based INR. For a given image coordinate, the model retrieves the K nearest Gaussians, aggregates distance-weighted embeddings, and predicts the RGB value via a neural network. This design enables continuous image representation, interpretable geometric structure, and flexible local editing, providing a foundation for physically aware and interactive image manipulation. Our method supports geometry-consistent transformations, seamless super-resolution, and integration with physics-based simulations. Moreover, the Gaussian representation allows lifting a single 2D image into a geometry-aware 3D representation, enabling depth-guided editing. Experiments demonstrate that GaINeR achieves state-of-the-art reconstruction quality while maintaining flexible and physically consistent image editing. The official implementation of our method is publicly available at https://github.com/WJakubowska/GaINeR.