Diffusion in the Dark: A Diffusion Model for Low-Light Text Recognition

TL;DR

This paper introduces Diffusion in the Dark (DiD), a diffusion model that reconstructs low-light images with high-frequency detail preservation, significantly improving text recognition in dark conditions without task-specific tuning.

Contribution

The paper presents a novel diffusion-based approach for low-light image reconstruction that outperforms existing methods in low-light text recognition tasks.

Findings

DiD achieves qualitatively competitive reconstructions.

DiD outperforms state-of-the-art low-light methods in text recognition.

DiD preserves high-frequency details even in extremely noisy conditions.

Abstract

Capturing images is a key part of automation for high-level tasks such as scene text recognition. Low-light conditions pose a challenge for high-level perception stacks, which are often optimized on well-lit, artifact-free images. Reconstruction methods for low-light images can produce well-lit counterparts, but typically at the cost of high-frequency details critical for downstream tasks. We propose Diffusion in the Dark (DiD), a diffusion model for low-light image reconstruction for text recognition. DiD provides qualitatively competitive reconstructions with that of state-of-the-art (SOTA), while preserving high-frequency details even in extremely noisy, dark conditions. We demonstrate that DiD, without any task-specific optimization, can outperform SOTA low-light methods in low-light text recognition on real images, bolstering the potential of diffusion models to solve ill-posed inverse problems.