Video Deblurring by Sharpness Prior Detection and Edge Information

2025-01-21

AI Summary

Existing video deblurring methods exhibit limited performance in critical applications such as autonomous driving and face recognition: conventional approaches rely solely on motion modeling, while mainstream video-based methods suffer from insufficient generalization and robustness due to their dependence on a fixed number of sharp-frame priors. To address these limitations, we propose SPEINet, an end-to-end deblurring framework incorporating variable-frequency sharp-frame priors and edge guidance. SPEINet features a novel attention-based encoder-decoder architecture, a lightweight tunable sharp-frame detection module, and an edge-feature extraction module. We further introduce GoProRS—the first benchmark dataset supporting controllable sharp-frame frequency. Trained via self-supervised and weakly supervised strategies, SPEINet achieves an average PSNR gain of 3.2 dB across multiple benchmarks, demonstrating significantly improved cross-domain generalization and robustness to real-world blur.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Learning on the Edge & Model Compression
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Responsible Web: Data and user privacy-enhancing technologies for the Web
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Cloud, edge and content delivery systems for the Web

Abstract

Video deblurring is essential task for autonomous driving, facial recognition, and security surveillance. Traditional methods directly estimate motion blur kernels, often introducing artifacts and leading to poor results. Recent approaches utilize the detection of sharp frames within video sequences to enhance deblurring. However, existing datasets rely on fixed number of sharp frames, which may be too restrictive for some applications and may introduce a bias during model training. To address these limitations and enhance domain adaptability, this work first introduces GoPro Random Sharp (GoProRS), a new dataset where the the frequency of sharp frames within the sequence is customizable, allowing more diverse training and testing scenarios. Furthermore, it presents a novel video deblurring model, called SPEINet, that integrates sharp frame features into blurry frame reconstruction through an attention-based encoder-decoder architecture, a lightweight yet robust sharp frame detection and an edge extraction phase. Extensive experimental results demonstrate that SPEINet outperforms state-of-the-art methods across multiple datasets, achieving an average of +3.2% PSNR improvement over recent techniques. Given such promising results, we believe that both the proposed model and dataset pave the way for future advancements in video deblurring based on the detection of sharp frames.

Problem

• Video Deblurring
• Autonomous Vehicles
• Facial Recognition

Innovation

• GoProRS Dataset
• SPEINet Model
• Customizable Sharp Frames