BP-EVD: Forward Block-Output Propagation for Efficient Video Denoising

Ostrowski, Piotr K.; Katsaros, Efklidis; Węsierski, Daniel; Jezierska, Anna

doi:10.1109/tip.2022.3176210

Cited by 9 publications

(6 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, the processed video needs to be temporally consistent. These constraints shape already the very few methods dedicated to raw video denoising, which either resort to recurrent techniques [13,1,40,24,44], or limit themselves to small temporal windows of a few frames [57,64,53,63,59,7,54,33].…”

Section: Mf2fmentioning

confidence: 99%

“…For a video restoration task, it is impractical to consider a large window of input frames, which makes recurrent networks an appealing choice for integrating temporal information across a larger number frames beyond the input window. Recurrent networks have been applied to video denoising [40,24,44] and super-resolution [50,27,16]. To address for the first time the video JDD problem, we define a simple architecture that combines recurrence on the output frame [50] and feature recurrence [27,16,24].…”

Section: Recurrent Cnn For Video Jddmentioning

confidence: 99%

“…A diagram of the proposed Recurrent Video joint Denoising and Demosaicing (RVDD) method is given in Figure 2. We consider a standard U-Net CNN (similarly to [57,53,44,63]), which we denote by F, that receives four inputs: the previous RGB output u t−1 , the current and next raw noisy frames f t , f t+1 , and the feature map from the last hidden layer φ L t−1 of the previous frame (with C channels and spatial resolution W × H). The raw inputs f t and f t+1 are demosaiced with the Hamilton-Adams method [30], which we denote by D. The adjacent frames and activation maps are aligned to frame t using warping operators W t−1,t and W t+1,t to compensate for motion:…”

Section: Recurrent Cnn For Video Jddmentioning

confidence: 99%

See 2 more Smart Citations

Video joint denoising and demosaicing with recurrent CNNs

Dewil

Courtois

Rodríguez

et al. 2023

2023 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)

View full text Add to dashboard Cite

HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L'archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d'enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

show abstract

Section: Mf2fmentioning

confidence: 99%

Section: Recurrent Cnn For Video Jddmentioning

confidence: 99%

Section: Recurrent Cnn For Video Jddmentioning

confidence: 99%

See 1 more Smart Citation

Video joint denoising and demosaicing with recurrent CNNs

Dewil

Courtois

Rodríguez

et al. 2023

2023 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)

View full text Add to dashboard Cite

show abstract

“…The results of EDVR and FastDVDnet are quoted from [27]. The results of FastDVDnet-S are quoted from [66]. RVRT can only process 3-channel RGB inputs and can not process raw data, we apply a simple ISP to the raw inputs before feeding them to the optical flow network.…”

Section: Tablementioning

confidence: 99%

“…Therefore, we directly quote their scores from their papers. We further introduce BP-EVD [66] and LLRVD [14] for comparison. Since they are not open-sourced, we did not compare them on Re-CRVD dataset.…”

Section: Gmacsmentioning

confidence: 99%

Supervised Raw Video Denoising With a Benchmark Dataset on Dynamic Scenes

Yue

Cao

Liao

et al. 2020

2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

102

113

View full text Add to dashboard Cite

In recent years, raw video denoising has garnered increased attention due to the consistency with the imaging process and well-studied noise modeling in the raw domain. Despite these advancements, two problems still hinder the denoising performance. Firstly, there is no large dataset with realistic motions for supervised raw video denoising, as capturing noisy and clean frames for real dynamic scenes is difficult. To address this, we propose recapturing existing high-resolution videos displayed on a 4K screen. Specifically, we recapture the screen content with high-low ISO settings to construct noisy-clean paired frames. Afterward, we introduce intensity, spatial, and color correction strategies to make the paired frames well-aligned. Then, the aligned frames are concatenated with temporal order to construct paired videos. In this way, we construct a video denoising dataset (named as ReCRVD) with 120 groups of noisy-clean videos, whose ISO values ranging from 1600 to 25600. Secondly, while non-local temporal-spatial attention is beneficial for denoising, it often leads to heavy computation costs. In this work, we propose an efficient raw video denoising transformer network (RViDeformer) that explores both short and long-distance correlations. Specifically, we introduce Low-Resolution-Window Self-Attention (LWSA), Global-Window Self-Attention (GWSA), and Neighbour-Window Self-Attention (NWSA) to build a multi-branch spatial self-attention for spatial reconstruction. Similarly, Global-Window Temporal Mutual Attention (GTMA) and Neighbour-Window Temporal Mutual Attention (NTMA) are proposed to build multi-branch temporal self-attention for temporal reconstruction. We employ reparameterization to reduce computation costs. Our network is trained in both supervised and unsupervised manners, achieving the best performance compared with state-of-the-art methods. Additionaly, the model trained with our proposed dataset (ReCRVD) outperforms the model trained with previous benchmark dataset (CRVD) when evaluated on the real-world outdoor noisy videos. Our code and dataset will be released after the acceptance of this work.

show abstract