Enregistré dans:
Détails bibliographiques
Auteurs principaux: Ji, Bo, Yao, Angela
Format: Preprint
Publié: 2024
Sujets:
Accès en ligne:https://arxiv.org/abs/2412.01559
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866916503015653376
author Ji, Bo
Yao, Angela
author_facet Ji, Bo
Yao, Angela
contents State-of-the-art video deblurring methods use deep network architectures to recover sharpened video frames. Blurring especially degrades high-frequency (HF) information, yet this aspect is often overlooked by recent models that focus more on enhancing architectural design. Recovering these fine details is challenging, partly due to the spectral bias of neural networks, which are inclined towards learning low-frequency functions. To address this, we enforce explicit network structures to capture the fine details and edges. We dynamically predict adaptive high-pass kernels from a linear combination of high-pass basis kernels to extract high-frequency features. This strategy is highly efficient, resulting in low-memory footprints for training and fast run times for inference, all while achieving state-of-the-art when compared to low-budget models. The code is available at https://github.com/jibo27/AHFNet.
format Preprint
id arxiv_https___arxiv_org_abs_2412_01559
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Adaptive High-Pass Kernel Prediction for Efficient Video Deblurring
Ji, Bo
Yao, Angela
Computer Vision and Pattern Recognition
State-of-the-art video deblurring methods use deep network architectures to recover sharpened video frames. Blurring especially degrades high-frequency (HF) information, yet this aspect is often overlooked by recent models that focus more on enhancing architectural design. Recovering these fine details is challenging, partly due to the spectral bias of neural networks, which are inclined towards learning low-frequency functions. To address this, we enforce explicit network structures to capture the fine details and edges. We dynamically predict adaptive high-pass kernels from a linear combination of high-pass basis kernels to extract high-frequency features. This strategy is highly efficient, resulting in low-memory footprints for training and fast run times for inference, all while achieving state-of-the-art when compared to low-budget models. The code is available at https://github.com/jibo27/AHFNet.
title Adaptive High-Pass Kernel Prediction for Efficient Video Deblurring
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2412.01559