Bidirectional Motion Estimation with Cyclic Cost Volume for High Dynamic Range Imaging

Vien, An Gia; Park, Seonghyun; Nhat, Truong Thanh; Kim, Gahyeon; Lee, Chul

doi:10.1109/cvprw56347.2022.00125

Cited by 6 publications

(6 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Additionally, all the methods were close in SSIM, however, our method was able to outperform the SOTA in both SSIM and µ − SSIM . Furthermore, although our result with the value of 0.03 is the second best in LPIPS, it performed worst in delta-E. On the other hand, Vien et al [33] had the lowest GMACs value, and GSANet is ranked second lowest. Moreover, it is visible that in terms of the number of parameters, GSANet has the lowest and the proposed method is in the second place among the algorithms.…”

Section: Evaluation Metrics and Comparison 1) Quantitative Comparisoncontrasting

confidence: 52%

“…More specifically, Fig. 11 demonstrates the results of ours, DRHDR [26], Vien et al [33], and GSANet [24]. As can be seen, the output of Vien et al in the first scene has distortion in the bright areas, and it is visible that the algorithm cannot restore the details from these areas correctly.…”

Section: ) Qualitative Comparisonmentioning

confidence: 82%

“…Finally, they merged the information to produce an HDR image without ghosting. In [33], a model based on bidirectional motion estimation was proposed, in which the amount of optical flow between LDR images was estimated by motion estimation with cyclic cost volume and spatial attention maps, and eventually, an HDR image was produced with the help of the extracted local and global features. [34] implemented the first multi-bracket HDR pipeline using event cameras, in which they merged the extracted features of images and the events to produce an HDR image.…”

Section: A Multi-exposure Methodsmentioning

confidence: 99%

“…In the latter track, it was asked to reduce the GM ACs value to less than the baseline method while the P SN R and µ − P SN R values are almost the same as the baseline method. The proposed method has been compared with GSANet [24], DRHDR [26], and Vein et al [33] methods. As can be seen, Table 2 shows the proposed method has the highest value in terms of P SN R, while having the second highest value in µ − P SN R.…”

Section: Evaluation Metrics and Comparison 1) Quantitative Comparisonmentioning

confidence: 99%

See 3 more Smart Citations

High Dynamic Range Imaging via Visual Attention Modules

Omrani,

Moroni

2024

IEEE Access

View full text Add to dashboard Cite

Thanks to High Dynamic Range (HDR) imaging methods, the scope of photography has seen profound changes recently. To be more specific, such methods try to reconstruct the lost luminosity of the real world caused by the limitation of regular cameras from the Low Dynamic Range (LDR) images. Additionally, although the State-Of-The-Art (SOTA) methods in this topic perform well, they mainly concentrate on combining different exposures and pay less attention to extracting the informative parts of the images. Thus, this paper aims to introduce a new model capable of incorporating information from the most visible areas of each image extracted by a Visual Attention Module (VAM) which is a result of a segmentation strategy.In particular, the model, based on a deep learning architecture, utilizes the extracted areas to produce the final HDR image. The results demonstrate that our method outperformed most of the SOTA algorithms.

show abstract

Section: Evaluation Metrics and Comparison 1) Quantitative Comparisoncontrasting

confidence: 52%

Section: ) Qualitative Comparisonmentioning

confidence: 82%

Section: A Multi-exposure Methodsmentioning

confidence: 99%

Section: Evaluation Metrics and Comparison 1) Quantitative Comparisonmentioning

confidence: 99%

See 2 more Smart Citations

High Dynamic Range Imaging via Visual Attention Modules

Omrani,

Moroni

2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…One of the popular methods for dealing with this issue is aligning multiple LDR images explicitly as a preprocessing step and merging them with conventional methods for static scenes. For example, homography transformation and optical flow are widely used as aligning methods [3], [4], [12], [18], [23]. However, the homography transformation cannot align foreground moving objects, and optical flow generally propagates estimation error to the reconstruction stage.…”

mentioning

confidence: 99%

Frequency-Domain Multi-Exposure HDR Imaging Network With Representative Image Features

Lee,

Park,

Jang

et al. 2023

IEEE Access

View full text Add to dashboard Cite

Constructing a high dynamic range (HDR) image from multi-exposure low dynamic range (LDR) images is challenging mainly due to two major problems. One is a large misalignment between the LDR images taken at different moments due to camera and object motions. The other is missing content in over-or under-exposed areas, which can be very large in the LDR images with long or short exposure times. These problems lead to ghosting artifacts or saturated regions in the reconstructed HDR image. In this paper, we propose a method using convolutional neural networks (CNNs) to address these problems. Specifically, each LDR image is fed to a two-branch CNN that extracts two kinds of features separately, named content features and global features. Then, another CNN aligns and merges the features to create the HDR output. The content features are used to create spatial attention masks, which are used to align the features in the time domain with exposure-invariant attributes. Further, Fourier coefficient vectors of global features are utilized to modulate intermediate features in the frequency domain during the reconstruction process. Experimental results show that our method achieves state-of-the-art HDR reconstruction performances on several benchmarks. Our code is available at https://github.com/KeuntekLee/FDM-HDR. INDEX TERMSConvolutional neural networks, frequency domain feature modulating, high dynamic range imaging, multi-exposure fusion I. INTRODUCTION Recently, deep learning-based methods have shown impressive performances in diverse domains. Especially, convolutional neural networks (CNNs) have achieved notable

show abstract

Ghost-Free High Dynamic Range Imaging Based on Two-Stage Dense Image Alignment

Demirbilek,

Çiftçi

2024

Engineering Cyber-Physical Systems and Critical Infrastructures

View full text Add to dashboard Cite

Bidirectional Motion Estimation with Cyclic Cost Volume for High Dynamic Range Imaging

Cited by 6 publications

References 24 publications

High Dynamic Range Imaging via Visual Attention Modules

High Dynamic Range Imaging via Visual Attention Modules

Frequency-Domain Multi-Exposure HDR Imaging Network With Representative Image Features

Ghost-Free High Dynamic Range Imaging Based on Two-Stage Dense Image Alignment

Contact Info

Product

Resources

About