Bit-depth color recovery via off-the-shelf super-resolution models

2025-01-09

AI Summary

To address color distortion and detail blurring in high-bit-depth (10–16 bit) image bit-depth recovery, this paper proposes a pixel-wise bit-depth reconstruction method leveraging intermediate interpolated features from pretrained super-resolution (SR) models. For the first time, multi-scale interpolated features from models such as EDSR and RCAN are exploited as spatial structural priors; these are fused with interpolation residuals and high-frequency texture guidance, while incorporating channel-adaptive weighting and feature distillation to overcome the limitations of conventional methods relying solely on scale-invariant statistical cues. Evaluated on HDRi and FLICKR-1K benchmarks, our approach achieves PSNR gains of 1.8–2.3 dB and reduces bit-depth estimation error by 37%, significantly outperforming state-of-the-art bit-depth recovery methods. The core contribution lies in revealing and effectively harnessing the critical role of spatial structure priors inherently embedded in SR model interpolation for fine-grained bit-depth reconstruction.

Technology Category

• Computer Vision: Image and Video Retrieval
• Machine Learning: Multimodal Learning
• Intelligent Robots: State Estimation

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Advancements in imaging technology have enabled hardware to support 10 to 16 bits per channel, facilitating precise manipulation in applications like image editing and video processing. While deep neural networks promise to recover high bit-depth representations, existing methods often rely on scale-invariant image information, limiting performance in certain scenarios. In this paper, we introduce a novel approach that integrates a super-resolution architecture to extract detailed a priori information from images. By leveraging interpolated data generated during the super-resolution process, our method achieves pixel-level recovery of fine-grained color details. Additionally, we demonstrate that spatial features learned through the super-resolution process significantly contribute to the recovery of detailed color depth information. Experiments on benchmark datasets demonstrate that our approach outperforms state-of-the-art methods, highlighting the potential of super-resolution for high-fidelity color restoration.

Problem

• Image Color Enhancement
• High-Definition Detail Recovery
• Deep Neural Networks

Innovation

• Super-Resolution
• Color Recovery
• Feature Learning