Learning a Deep Convolutional Network for Light-Field Image Super-Resolution

Youngjin Yoon, Hae-Gon Jeon, Donggeun Yoo, Joon-Young Lee, In So Kweon; Proceedings of the IEEE International Conference on Computer Vision (ICCV) Workshops, 2015, pp. 24-32

Abstract


Commercial Light-Field cameras provide spatial and angular information, but its limited resolution becomes an important problem in practical use. In this paper, we present a novel method for Light-Field image super-resolution (SR) via a deep convolutional neural network. Rather than the conventional optimization framework, we adopt a datadriven learning method to simultaneously up-sample the angular resolution as well as the spatial resolution of a Light-Field image. We first augment the spatial resolution of each sub-aperture image to enhance details by a spatial SR network. Then, novel views between the sub-aperture images are generated by an angular super-resolution network. These networks are trained independently but finally finetuned via end-to-end training. The proposed method shows the state-of-the-art performance on HCI synthetic dataset, and is further evaluated by challenging real-world applications including refocusing and depth map estimation.

Related Material


[pdf]
[bibtex]
@InProceedings{Yoon_2015_ICCV_Workshops,
author = {Yoon, Youngjin and Jeon, Hae-Gon and Yoo, Donggeun and Lee, Joon-Young and So Kweon, In},
title = {Learning a Deep Convolutional Network for Light-Field Image Super-Resolution},
booktitle = {Proceedings of the IEEE International Conference on Computer Vision (ICCV) Workshops},
month = {December},
year = {2015}
}