ETRI-Knowledge Sharing Plaform

ENGLISH

성과물

논문 검색
구분 SCI
연도 ~ 키워드

상세정보

학술지 Enhanced Single Image Super Resolution Method Using Lightweight Multi-Scale Channel Dense Network
Cited 3 time in scopus Download 11 time Share share facebook twitter linkedin kakaostory
저자
이유호, 전동산, 김병규, 이헌주
발행일
202105
출처
Sensors, v.21 no.10, pp.1-17
ISSN
1424-8220
출판사
MDPI
DOI
https://dx.doi.org/10.3390/s21103351
협약과제
21IR1200, 스마트 도로조명 정보처리·통신 플랫폼 개발, 이헌주
초록
Super resolution (SR) enables to generate a high-resolution (HR) image from one or more low-resolution (LR) images. Since a variety of CNN models have been recently studied in the areas of computer vision, these approaches have been combined with SR in order to provide higher image restoration. In this paper, we propose a lightweight CNN-based SR method, named multiscale channel dense network (MCDN). In order to design the proposed network, we extracted the training images from the DIVerse 2K (DIV2K) dataset and investigated the trade-off between the SR accuracy and the network complexity. The experimental results show that the proposed method can significantly reduce the network complexity, such as the number of network parameters and total memory capacity, while maintaining slightly better or similar perceptual quality compared to the previous methods.
키워드
Convolutional neural network, Deep learning, Lightweight neural network, Super resolution
KSP 제안 키워드
Computer Vision(CV), Convolution neural network(CNN), Dense network, High-resolution, Image restoration, Multi-scale, Network Parameters, Perceptual Quality, Resolution method, Trade-off, deep learning(DL)
본 저작물은 크리에이티브 커먼즈 저작자 표시 (CC BY) 조건에 따라 이용할 수 있습니다.
저작자 표시 (CC BY)