Efficient Deep Models for Real-Time 4K Image Super-Resolution. NTIRE 2023 Benchmark and Report

Marcos V. Conde; Eduard Zamfir; Radu Timofte; Daniel Motilla; Cen Liu; Zexin Zhang; Yunbo Peng; Yue Lin; Jiaming Guo; Xueyi Zou; Yuyi Chen; Yi Liu; Jia Hao; Youliang Yan; Yuanfan Zhang; Gen Li; Lei Sun; Lingshun Kong; Haoran Bai; Jinshan Pan; Jiangxin Dong; Jinhui Tang; Mustafa Ayazoglu; Bahri Batuhan Bilecen; Mingxi Li; Yuhang Zhang; Xianjun Fan; Yankai Sheng; Long Sun; Zibin Liu; Weiran Gou; Shaoqing Li; Ziyao Yi; Yan Xiang; Dehui Kong; Ke Xu; Ganzorig Gankhuyag; Kihwan Yoon; Jin Zhang; Gaocheng Yu; Feng Zhang; Hongbin Wang; Zhou Zhou; Jiahao Chao; Hongfan Gao; Jiali Gong; Zhengfeng Yang; Zhenbing Zeng; Chengpeng Chen; Zichao Guo; Anjin Park; Yuqing Liu; Qi Jia; Hongyuan Yu; Xuanwu Yin; Kunlong Zuo; Dongyang Zhang; Ting Fu; Zhengxue Cheng; Shiai Zhu; Dajiang Zhou; Weichen Yu; Lin Ge; Jiahua Dong; Yajun Zou; Zhuoyuan Wu; Binnan Han; Xiaolin Zhang; Heng Zhang; Ben Shao; Shaolong Zheng; Daheng Yin; Baijun Chen; Mengyang Liu; Marian Sergiu Nistor; Yi Chung Chen; Zhi Kai Huang; Yuan Chun Chiang; Wei Ting Chen; Hao Hsiang Yang; Hua En Chang; I. Hsiang Chen; Chia Hsuan Hsieh; Sy Yen Kuo; Tu Vo; Qingsen Yan; Yun Zhu; Jinqiu Su; Yanning Zhang; Cheng Zhang; Jiaying Luo; Youngsun Cho; Nakyung Lee

doi:10.1109/CVPRW59228.2023.00154

Efficient Deep Models for Real-Time 4K Image Super-Resolution. NTIRE 2023 Benchmark and Report

Marcos V. Conde, Eduard Zamfir, Radu Timofte, Daniel Motilla, Cen Liu, Zexin Zhang, Yunbo Peng, Yue Lin, Jiaming Guo, Xueyi Zou, Yuyi Chen, Yi Liu, Jia Hao, Youliang Yan, Yuanfan Zhang, Gen Li, Lei Sun, Lingshun Kong, Haoran Bai, Jinshan PanJiangxin Dong, Jinhui Tang, Mustafa Ayazoglu, Bahri Batuhan Bilecen, Mingxi Li, Yuhang Zhang, Xianjun Fan, Yankai Sheng, Long Sun, Zibin Liu, Weiran Gou, Shaoqing Li, Ziyao Yi, Yan Xiang, Dehui Kong, Ke Xu, Ganzorig Gankhuyag, Kihwan Yoon, Jin Zhang, Gaocheng Yu, Feng Zhang, Hongbin Wang, Zhou Zhou, Jiahao Chao, Hongfan Gao, Jiali Gong, Zhengfeng Yang, Zhenbing Zeng, Chengpeng Chen, Zichao Guo, Anjin Park, Yuqing Liu, Qi Jia, Hongyuan Yu, Xuanwu Yin, Kunlong Zuo, Dongyang Zhang, Ting Fu, Zhengxue Cheng, Shiai Zhu, Dajiang Zhou, Weichen Yu, Lin Ge, Jiahua Dong, Yajun Zou, Zhuoyuan Wu, Binnan Han, Xiaolin Zhang, Heng Zhang, Ben Shao, Shaolong Zheng, Daheng Yin, Baijun Chen, Mengyang Liu, Marian Sergiu Nistor, Yi Chung Chen, Zhi Kai Huang, Yuan Chun Chiang, Wei Ting Chen, Hao Hsiang Yang, Hua En Chang, I. Hsiang Chen, Chia Hsuan Hsieh, Sy Yen Kuo, Tu Vo, Qingsen Yan, Yun Zhu, Jinqiu Su, Yanning Zhang, Cheng Zhang, Jiaying Luo, Youngsun Cho, Nakyung Lee

计算机学院

科研成果: 书/报告/会议事项章节 › 会议稿件 › 同行评审

25 引用（Scopus）

摘要

This paper introduces a novel benchmark for efficient up-scaling as part of the NTIRE 2023 Real-Time Image Super-Resolution (RTSR) Challenge, which aimed to upscale images from 720p and 1080p resolution to native 4K (×2 and ×3 factors) in real-time on commercial GPUs. For this, we use a new test set containing diverse 4K images ranging from digital art to gaming and photography. We assessed the methods devised for 4K SR by measuring their runtime, parameters, and FLOPs, while ensuring a minimum PSNR fidelity over Bicubic interpolation. Out of the 170 participants, 25 teams contributed to this report, making it the most comprehensive benchmark to date and showcasing the latest advancements in real-time SR.

源语言	英语
主期刊名	Proceedings - 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2023
出版商	IEEE Computer Society
页	1495-1521
页数	27
ISBN（电子版）	9798350302493
DOI	https://doi.org/10.1109/CVPRW59228.2023.00154
出版状态	已出版 - 2023
活动	2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2023 - Vancouver, 加拿大期限: 18 6月 2023 → 22 6月 2023

出版系列

姓名	IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops
卷	2023-June
ISSN（印刷版）	2160-7508
ISSN（电子版）	2160-7516

会议

会议	2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2023
国家/地区	加拿大
市	Vancouver
时期	18/06/23 → 22/06/23

访问文件

10.1109/CVPRW59228.2023.00154

其它文件与链接

链接到 Scopus 的出版物

引用此

Conde, M. V., Zamfir, E., Timofte, R., Motilla, D., Liu, C., Zhang, Z., Peng, Y., Lin, Y., Guo, J., Zou, X., Chen, Y., Liu, Y., Hao, J., Yan, Y., Zhang, Y., Li, G., Sun, L., Kong, L., Bai, H., ... Lee, N. (2023). Efficient Deep Models for Real-Time 4K Image Super-Resolution. NTIRE 2023 Benchmark and Report. 在 Proceedings - 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2023 (页码 1495-1521). (IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops; 卷 2023-June). IEEE Computer Society. https://doi.org/10.1109/CVPRW59228.2023.00154

Conde, Marcos V. ; Zamfir, Eduard ; Timofte, Radu 等. / Efficient Deep Models for Real-Time 4K Image Super-Resolution. NTIRE 2023 Benchmark and Report. Proceedings - 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2023. IEEE Computer Society, 2023. 页码 1495-1521 (IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops).

@inproceedings{f7a3b08f8a594b8388bca9fa60144b73,

title = "Efficient Deep Models for Real-Time 4K Image Super-Resolution. NTIRE 2023 Benchmark and Report",

abstract = "This paper introduces a novel benchmark for efficient up-scaling as part of the NTIRE 2023 Real-Time Image Super-Resolution (RTSR) Challenge, which aimed to upscale images from 720p and 1080p resolution to native 4K (×2 and ×3 factors) in real-time on commercial GPUs. For this, we use a new test set containing diverse 4K images ranging from digital art to gaming and photography. We assessed the methods devised for 4K SR by measuring their runtime, parameters, and FLOPs, while ensuring a minimum PSNR fidelity over Bicubic interpolation. Out of the 170 participants, 25 teams contributed to this report, making it the most comprehensive benchmark to date and showcasing the latest advancements in real-time SR.",

author = "Conde, {Marcos V.} and Eduard Zamfir and Radu Timofte and Daniel Motilla and Cen Liu and Zexin Zhang and Yunbo Peng and Yue Lin and Jiaming Guo and Xueyi Zou and Yuyi Chen and Yi Liu and Jia Hao and Youliang Yan and Yuanfan Zhang and Gen Li and Lei Sun and Lingshun Kong and Haoran Bai and Jinshan Pan and Jiangxin Dong and Jinhui Tang and Mustafa Ayazoglu and Bilecen, {Bahri Batuhan} and Mingxi Li and Yuhang Zhang and Xianjun Fan and Yankai Sheng and Long Sun and Zibin Liu and Weiran Gou and Shaoqing Li and Ziyao Yi and Yan Xiang and Dehui Kong and Ke Xu and Ganzorig Gankhuyag and Kihwan Yoon and Jin Zhang and Gaocheng Yu and Feng Zhang and Hongbin Wang and Zhou Zhou and Jiahao Chao and Hongfan Gao and Jiali Gong and Zhengfeng Yang and Zhenbing Zeng and Chengpeng Chen and Zichao Guo and Anjin Park and Yuqing Liu and Qi Jia and Hongyuan Yu and Xuanwu Yin and Kunlong Zuo and Dongyang Zhang and Ting Fu and Zhengxue Cheng and Shiai Zhu and Dajiang Zhou and Weichen Yu and Lin Ge and Jiahua Dong and Yajun Zou and Zhuoyuan Wu and Binnan Han and Xiaolin Zhang and Heng Zhang and Ben Shao and Shaolong Zheng and Daheng Yin and Baijun Chen and Mengyang Liu and Nistor, {Marian Sergiu} and Chen, {Yi Chung} and Huang, {Zhi Kai} and Chiang, {Yuan Chun} and Chen, {Wei Ting} and Yang, {Hao Hsiang} and Chang, {Hua En} and Chen, {I. Hsiang} and Hsieh, {Chia Hsuan} and Kuo, {Sy Yen} and Tu Vo and Qingsen Yan and Yun Zhu and Jinqiu Su and Yanning Zhang and Cheng Zhang and Jiaying Luo and Youngsun Cho and Nakyung Lee",

note = "Publisher Copyright: {\textcopyright} 2023 IEEE.; 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2023 ; Conference date: 18-06-2023 Through 22-06-2023",

year = "2023",

doi = "10.1109/CVPRW59228.2023.00154",

language = "英语",

series = "IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops",

publisher = "IEEE Computer Society",

pages = "1495--1521",

booktitle = "Proceedings - 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2023",

}

Conde, MV, Zamfir, E, Timofte, R, Motilla, D, Liu, C, Zhang, Z, Peng, Y, Lin, Y, Guo, J, Zou, X, Chen, Y, Liu, Y, Hao, J, Yan, Y, Zhang, Y, Li, G, Sun, L, Kong, L, Bai, H, Pan, J, Dong, J, Tang, J, Ayazoglu, M, Bilecen, BB, Li, M, Zhang, Y, Fan, X, Sheng, Y, Sun, L, Liu, Z, Gou, W, Li, S, Yi, Z, Xiang, Y, Kong, D, Xu, K, Gankhuyag, G, Yoon, K, Zhang, J, Yu, G, Zhang, F, Wang, H, Zhou, Z, Chao, J, Gao, H, Gong, J, Yang, Z, Zeng, Z, Chen, C, Guo, Z, Park, A, Liu, Y, Jia, Q, Yu, H, Yin, X, Zuo, K, Zhang, D, Fu, T, Cheng, Z, Zhu, S, Zhou, D, Yu, W, Ge, L, Dong, J, Zou, Y, Wu, Z, Han, B, Zhang, X, Zhang, H, Shao, B, Zheng, S, Yin, D, Chen, B, Liu, M, Nistor, MS, Chen, YC, Huang, ZK, Chiang, YC, Chen, WT, Yang, HH, Chang, HE, Chen, IH, Hsieh, CH, Kuo, SY, Vo, T, Yan, Q, Zhu, Y, Su, J, Zhang, Y, Zhang, C, Luo, J, Cho, Y & Lee, N 2023, Efficient Deep Models for Real-Time 4K Image Super-Resolution. NTIRE 2023 Benchmark and Report. 在 Proceedings - 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2023. IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, 卷 2023-June, IEEE Computer Society, 页码 1495-1521, 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2023, Vancouver, 加拿大, 18/06/23. https://doi.org/10.1109/CVPRW59228.2023.00154

Efficient Deep Models for Real-Time 4K Image Super-Resolution. NTIRE 2023 Benchmark and Report. / Conde, Marcos V.; Zamfir, Eduard; Timofte, Radu 等.
Proceedings - 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2023. IEEE Computer Society, 2023. 页码 1495-1521 (IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops; 卷 2023-June).

科研成果: 书/报告/会议事项章节 › 会议稿件 › 同行评审

TY - GEN

T1 - Efficient Deep Models for Real-Time 4K Image Super-Resolution. NTIRE 2023 Benchmark and Report

AU - Conde, Marcos V.

AU - Zamfir, Eduard

AU - Timofte, Radu

AU - Motilla, Daniel

AU - Liu, Cen

AU - Zhang, Zexin

AU - Peng, Yunbo

AU - Lin, Yue

AU - Guo, Jiaming

AU - Zou, Xueyi

AU - Chen, Yuyi

AU - Liu, Yi

AU - Hao, Jia

AU - Yan, Youliang

AU - Zhang, Yuanfan

AU - Li, Gen

AU - Sun, Lei

AU - Kong, Lingshun

AU - Bai, Haoran

AU - Pan, Jinshan

AU - Dong, Jiangxin

AU - Tang, Jinhui

AU - Ayazoglu, Mustafa

AU - Bilecen, Bahri Batuhan

AU - Li, Mingxi

AU - Zhang, Yuhang

AU - Fan, Xianjun

AU - Sheng, Yankai

AU - Sun, Long

AU - Liu, Zibin

AU - Gou, Weiran

AU - Li, Shaoqing

AU - Yi, Ziyao

AU - Xiang, Yan

AU - Kong, Dehui

AU - Xu, Ke

AU - Gankhuyag, Ganzorig

AU - Yoon, Kihwan

AU - Zhang, Jin

AU - Yu, Gaocheng

AU - Zhang, Feng

AU - Wang, Hongbin

AU - Zhou, Zhou

AU - Chao, Jiahao

AU - Gao, Hongfan

AU - Gong, Jiali

AU - Yang, Zhengfeng

AU - Zeng, Zhenbing

AU - Chen, Chengpeng

AU - Guo, Zichao

AU - Park, Anjin

AU - Liu, Yuqing

AU - Jia, Qi

AU - Yu, Hongyuan

AU - Yin, Xuanwu

AU - Zuo, Kunlong

AU - Zhang, Dongyang

AU - Fu, Ting

AU - Cheng, Zhengxue

AU - Zhu, Shiai

AU - Zhou, Dajiang

AU - Yu, Weichen

AU - Ge, Lin

AU - Dong, Jiahua

AU - Zou, Yajun

AU - Wu, Zhuoyuan

AU - Han, Binnan

AU - Zhang, Xiaolin

AU - Zhang, Heng

AU - Shao, Ben

AU - Zheng, Shaolong

AU - Yin, Daheng

AU - Chen, Baijun

AU - Liu, Mengyang

AU - Nistor, Marian Sergiu

AU - Chen, Yi Chung

AU - Huang, Zhi Kai

AU - Chiang, Yuan Chun

AU - Chen, Wei Ting

AU - Yang, Hao Hsiang

AU - Chang, Hua En

AU - Chen, I. Hsiang

AU - Hsieh, Chia Hsuan

AU - Kuo, Sy Yen

AU - Vo, Tu

AU - Yan, Qingsen

AU - Zhu, Yun

AU - Su, Jinqiu

AU - Zhang, Yanning

AU - Zhang, Cheng

AU - Luo, Jiaying

AU - Cho, Youngsun

AU - Lee, Nakyung

PY - 2023

Y1 - 2023

N2 - This paper introduces a novel benchmark for efficient up-scaling as part of the NTIRE 2023 Real-Time Image Super-Resolution (RTSR) Challenge, which aimed to upscale images from 720p and 1080p resolution to native 4K (×2 and ×3 factors) in real-time on commercial GPUs. For this, we use a new test set containing diverse 4K images ranging from digital art to gaming and photography. We assessed the methods devised for 4K SR by measuring their runtime, parameters, and FLOPs, while ensuring a minimum PSNR fidelity over Bicubic interpolation. Out of the 170 participants, 25 teams contributed to this report, making it the most comprehensive benchmark to date and showcasing the latest advancements in real-time SR.

AB - This paper introduces a novel benchmark for efficient up-scaling as part of the NTIRE 2023 Real-Time Image Super-Resolution (RTSR) Challenge, which aimed to upscale images from 720p and 1080p resolution to native 4K (×2 and ×3 factors) in real-time on commercial GPUs. For this, we use a new test set containing diverse 4K images ranging from digital art to gaming and photography. We assessed the methods devised for 4K SR by measuring their runtime, parameters, and FLOPs, while ensuring a minimum PSNR fidelity over Bicubic interpolation. Out of the 170 participants, 25 teams contributed to this report, making it the most comprehensive benchmark to date and showcasing the latest advancements in real-time SR.

UR - http://www.scopus.com/inward/record.url?scp=85161047882&partnerID=8YFLogxK

U2 - 10.1109/CVPRW59228.2023.00154

DO - 10.1109/CVPRW59228.2023.00154

M3 - 会议稿件

AN - SCOPUS:85161047882

T3 - IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops

SP - 1495

EP - 1521

BT - Proceedings - 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2023

PB - IEEE Computer Society

T2 - 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2023

Y2 - 18 June 2023 through 22 June 2023

ER -

Conde MV, Zamfir E, Timofte R, Motilla D, Liu C, Zhang Z 等. Efficient Deep Models for Real-Time 4K Image Super-Resolution. NTIRE 2023 Benchmark and Report. 在 Proceedings - 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2023. IEEE Computer Society. 2023. 页码 1495-1521. (IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops). doi: 10.1109/CVPRW59228.2023.00154

Efficient Deep Models for Real-Time 4K Image Super-Resolution. NTIRE 2023 Benchmark and Report

摘要

出版系列

会议

访问文件

其它文件与链接

指纹

引用此