While single-image super-resolution (SISR) has attracted substantial interest in recent years, the proposed approaches are limited to learning image priors in order to add high frequency details. In contrast, multi-frame super-resolution (MFSR) offers the possibility of reconstructing rich details by combining signal information from multiple shifted images. This key advantage, along with the increasing popularity of burst photography, have made MFSR an important problem for real-world applications.
We propose a novel architecture for the burst super-resolution task. Our network takes multiple noisy RAW images as input, and generates a denoised, super-resolved RGB image as output. This is achieved by explicitly aligning deep embeddings of the input frames using pixel-wise optical flow. The information from all frames are then adaptively merged using an attention-based fusion module. In order to enable training and evaluation on real-world data, we additionally introduce the BurstSR dataset, consisting of smartphone bursts and high-resolution DSLR ground-truth. We perform comprehensive experimental analysis, demonstrating the effectiveness of the proposed architecture.
Gated Multi-Resolution Transfer Network for Burst Restoration and Enhancement
Indian Institute of Technology Ropar(印度理工学院鲁尔基分校);Mohamed bin Zayed University of AI(穆罕默德·本·扎耶德人工智能大学);Inception Institute of AI(起源人工智能研究院);Australian National University(澳大利亚国立大学);Linköping University(林雪平大学)
BurstGP: Enhancing Raw Burst Image Super Resolution with Generative Priors
AI Center - Toronto, Samsung Electronics(多伦多AI中心、三星电子);University of Toronto(多伦多大学);Vector Institute(向量研究所);University Health Network(大学健康网络);York University(约克大学)
Lucas-Kanade Reloaded: End-to-End Super-Resolution from Raw Image Bursts
Inria(法国国家信息与自动化研究所);DIENS (ENS-PSL, CNRS, Inria)(DIENS(巴黎高等师范学院-PSL,法国国家科学研究中心,法国国家信息与自动化研究所));Univ. Grenoble Alpes(格勒诺布尔-阿尔卑斯大学);CNRS(法国国家科学研究中心);Grenoble INP(格勒诺布尔国立理工学院);LJK(LJK(让·库尔坦研究所));Center for Data Science, New York University(纽约大学数据科学中心)
High Dynamic Range and Super-Resolution from Raw Image Bursts
Inria(法国国家信息与自动化研究所);DI/ENS (ENS-PSL, CNRS, Inria)(巴黎高等师范学院计算机系(巴黎文理研究大学巴黎高师、法国国家科学研究中心、法国国家信息与自动化研究所));Université Paris-Saclay(巴黎萨克雷大学);ENS Paris-Saclay(巴黎-萨克雷高等师范学院);Centre Borelli(博雷利中心);Center for Data Science, New York University(纽约大学数据科学中心);Univ. Grenoble Alpes(格勒诺布尔阿尔卑斯大学);CNRS(法国国家科学研究中心);Grenoble INP(格勒诺布尔理工学院);LJK(让·昆茨曼实验室)
Deep learning image burst stacking to reconstruct high-resolution ground-based solar observations
University of Graz(格拉茨大学);National Center for Atmospheric Research(美国国家大气研究中心);Instituto de Astrofísica de Canarias(加那利天体物理研究所);Universidad de La Laguna(拉古纳大学);Max-Planck-Institut für Sonnensystemforschung(马克斯·普朗克太阳系研究所)