Beyond the Ground Truth: Enhanced Supervision for Image Restoration

2025-12-03

AI Summary

Supervised learning for real-world image restoration is hindered by the limited quality of ground-truth images in labeled datasets, primarily due to acquisition constraints. To address this, we propose a supervised quality enhancement framework that operates in the frequency domain. Specifically, we design an adaptive frequency mask to semantically align and mix the original ground truth with its super-resolved counterpart, yielding an enhanced ground truth rich in fine details and free from hallucinated artifacts. Furthermore, we introduce a conditional frequency mask generator and a lightweight output optimization network to jointly perform frequency-domain Mixup and model fine-tuning. Extensive experiments demonstrate substantial improvements in reconstruction quality across multiple image restoration models under real-world conditions. A user study further validates the effectiveness of both the supervision enhancement and the output optimization components.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Semi-Supervised Learning
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Deep learning-based image restoration has achieved significant success. However, when addressing real-world degradations, model performance is limited by the quality of ground-truth images in datasets due to practical constraints in data acquisition. To address this limitation, we propose a novel framework that enhances existing ground truth images to provide higher-quality supervision for real-world restoration. Our framework generates perceptually enhanced ground truth images using super-resolution by incorporating adaptive frequency masks, which are learned by a conditional frequency mask generator. These masks guide the optimal fusion of frequency components from the original ground truth and its super-resolved variants, yielding enhanced ground truth images. This frequency-domain mixup preserves the semantic consistency of the original content while selectively enriching perceptual details, preventing hallucinated artifacts that could compromise fidelity. The enhanced ground truth images are used to train a lightweight output refinement network that can be seamlessly integrated with existing restoration models. Extensive experiments demonstrate that our approach consistently improves the quality of restored images. We further validate the effectiveness of both supervision enhancement and output refinement through user studies. Code is available at https://github.com/dhryougit/Beyond-the-Ground-Truth.

Problem

• Enhances ground truth images for better supervision
• Uses adaptive frequency masks for perceptual detail enrichment
• Trains a refinement network to improve restoration models

Innovation

• Generates enhanced ground truth via adaptive frequency masks
• Uses frequency-domain mixup to enrich details without artifacts
• Trains lightweight refinement network for existing restoration models