Lightweight improved residual network for efficient inverse tone mapping

被引:0
作者
Xue, Liqi [1 ]
Xu, Tianyi [1 ]
Song, Yongbao [2 ]
Liu, Yan [1 ]
Zhang, Lei [3 ]
Zhen, Xiantong [3 ]
Xu, Jun [1 ,4 ]
机构
[1] Nankai Univ, Sch Stat & Data Sci, Tianjin 300071, Peoples R China
[2] Nankai Univ, Sch Math Sci, Tianjin 300071, Peoples R China
[3] Guangdong Univ Petrochem Technol, Comp Sci Coll, Maoming 525000, Peoples R China
[4] Chinese Univ Hong Kong Shenzhen, Guangdong Prov Key Lab Big Data Comp, Shenzhen 518172, Peoples R China
基金
中国国家自然科学基金;
关键词
Inverse tone mapping; Improved residual block; Lightweight network; Inference efficiency;
D O I
10.1007/s11042-023-17811-7
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
The display devices like HDR10 televisions are increasingly prevalent in our daily life for visualizing high dynamic range (HDR) images. But the majority of media images on the internet remain in 8-bit standard dynamic range (SDR) format. Therefore, converting SDR images to HDR ones by inverse tone mapping (ITM) is crucial to unlock the full potential of abundant media images. However, existing ITM methods are usually developed with complex network architectures requiring huge computational costs. In this paper, we propose a lightweight Improved Residual Network (IRNet) by enhancing the power of popular residual block for efficient ITM. Specifically, we propose a new Improved Residual Block (IRB) to extract and fuse multi-layer features for fine-grained HDR image reconstruction. Experiments on three benchmark datasets demonstrate that our IRNet achieves state-of-the-art performance on both the ITM and joint SR-ITM tasks. The code, models and data will be publicly available at https://github.com/ThisisVikki/ITM-baseline.
引用
收藏
页码:67059 / 67082
页数:24
相关论文
共 72 条
[1]   Learning Multi-Scale Photo Exposure Correction [J].
Afifi, Mahmoud ;
Derpanis, Konstantinos G. ;
Ommer, Bjoern ;
Brown, Michael S. .
2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2021, 2021, :9153-9163
[2]  
Akyüz AG, 2007, ACM T GRAPHIC, V26, DOI [10.1145/1239451.1239489, 10.1145/1276377.1276425]
[3]   Inverse tone mapping [J].
University of Bristol .
Proc. GRAPHITE Int. Conf. Comput. Graph. Interact. Techniq. Australasia and Southeast Asia, 2006, (349-356) :349-356
[4]  
Banterle F., 2008, P 24 SPRING C COMPUT, P33, DOI 10.1145/1921264.1921275
[5]   Compression Efficiency of High Dynamic Range and Wide Color Gamut Pixel's Representation [J].
Boitard, Ronan ;
Pourazad, Mahsa T. ;
Nasiopoulos, Panos .
IEEE TRANSACTIONS ON BROADCASTING, 2018, 64 (01) :1-10
[6]   THE LAPLACIAN PYRAMID AS A COMPACT IMAGE CODE [J].
BURT, PJ ;
ADELSON, EH .
IEEE TRANSACTIONS ON COMMUNICATIONS, 1983, 31 (04) :532-540
[7]   Bilateral Guided Upsampling [J].
Chen, Jiawen ;
Adams, Andrew ;
Wadhwa, Neal ;
Hasinoff, Samuel W. .
ACM TRANSACTIONS ON GRAPHICS, 2016, 35 (06)
[8]   Attention-Guided Progressive Neural Texture Fusion for High Dynamic Range Image Restoration [J].
Chen, Jie ;
Yang, Zaifeng ;
Chan, Tsz Nam ;
Li, Hui ;
Hou, Junhui ;
Chau, Lap-Pui .
IEEE TRANSACTIONS ON IMAGE PROCESSING, 2022, 31 :2661-2672
[9]  
Chen X., 2021, IEEE INT C COMPUT VI, P4500
[10]   HDRUNet: Single Image HDR Reconstruction with Denoising and Dequantization [J].
Chen, Xiangyu ;
Liu, Yihao ;
Zhang, Zhengwen ;
Qiao, Yu ;
Dong, Chao .
2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION WORKSHOPS, CVPRW 2021, 2021, :354-363