TITAN: Bringing The Deep Image Prior to Implicit Representations

TL;DR

TITAN enhances implicit neural representations for images by integrating deep image priors, significantly improving interpolation, super-resolution, and tomography results through a residual connection and sparse weight constraints.

Contribution

The paper introduces TITAN, a novel method that incorporates deep image priors into INRs, improving their natural image interpolation and reconstruction capabilities.

Findings

Significant improvement in image interpolation and super-resolution.

Enhanced image sharpness with sparse weight constraints.

Better performance in computed tomography reconstructions.

Abstract

We study the interpolation capabilities of implicit neural representations (INRs) of images. In principle, INRs promise a number of advantages, such as continuous derivatives and arbitrary sampling, being freed from the restrictions of a raster grid. However, empirically, INRs have been observed to poorly interpolate between the pixels of the fit image; in other words, they do not inherently possess a suitable prior for natural images. In this paper, we propose to address and improve INRs' interpolation capabilities by explicitly integrating image prior information into the INR architecture via deep decoder, a specific implementation of the deep image prior (DIP). Our method, which we call TITAN, leverages a residual connection from the input which enables integrating the principles of the grid-based DIP into the grid-free INR. Through super-resolution and computed tomography experiments, we demonstrate that our method significantly improves upon classic INRs, thanks to the induced natural image bias. We also find that by constraining the weights to be sparse, image quality and sharpness are enhanced, increasing the Lipschitz constant.