Light Field Image Sparse Coding via CNN-Based EPI Super-Resolution

Zhao, Jinbo; An, Ping; Shan, Liang; Ma, Ran

doi:10.1109/vcip.2018.8698714

Cited by 10 publications

(14 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Combining them together can achieve great performance of LFI compression. Next, we compare the proposed method with Zhao et al [7], and stable improvement is achieved, which proves that decoder-side quality enhancement and newly designed MSDRN play an important role.…”

Section: Effectiveness Of the Lfi Compression Frameworkmentioning

confidence: 65%

“…So we propose two CNN-based schemes to suppress the compression distortions and improve the quality of reconstructed dense SAIs. The first one is proposed in our previous work [7], but it does not perform so well when the bitrate is high. To remedy this drawback, a new scheme is employed.…”

Section: Decoder-side Sais Quality Enhancement a Problem Analysismentioning

confidence: 99%

“…The sequence is compressed with JEM software. 3) Zhao et al: Our previous work [7], in which part of the SAIs are arranged as a pseudo-video sequence and compressed by HEVC. CNN-based EPI super-resolution is used to predict those non-key SAIs.…”

Section: Effectiveness Of the Lfi Compression Frameworkmentioning

confidence: 99%

“…Compared with [7], we consider the influence of compression distortions and add a decoder-side quality enhancement network to improve the performance.…”

Section: Effectiveness Of the Lfi Compression Frameworkmentioning

confidence: 99%

“…The loss of SAIs quality at the decoding side is mainly due to the quantization in lossy compression, so if we can recover the distortions to some extent without increasing the amount of data of the stream, the quality of reconstructing SAIs will be improved, which will also indirectly lead to an increase in compression efficiency. This paper mainly focuses on this aspect compared to our previous work [7]. Via convolutional neural network (CNN), we can recover the distortions of decoded key SAIs without extra information.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Light Field Image Compression via CNN-Based EPI Super-Resolution and Decoder-Side Quality Enhancement

Zhao

Yang

et al. 2019

IEEE Access

Self Cite

View full text Add to dashboard Cite

Because of the capacity of capturing both the spatial and angular information of the light rays simultaneously, light field images (LFIs) contain richer scene information compared with conventional images, but at the cost of huge volume. This paper proposes a novel LFI sparse compression framework driven by convolutional neural network (CNN). The epipolar plane image (EPI) super-resolution is for compensating the information loss caused by sparse sampling and the decoder-side sub-aperture images (SAIs) quality enhancement is for compensating the information loss caused by lossy compression. Specifically, we choose those SAIs both in odd rows and odd columns as our key SAIs and compress them using standard video encoder. For those non-key SAIs, we predict them using decompressed key SAIs by taking advantage of the special structure of EPI. The low-resolution EPIs generated from the sparse SAIs are super-resolved by a CNN and the outputs, high-resolution EPIs, are used to rebuild the dense SAIs. Moreover, in order to improve the quality of the predicted SAI, we add decoder-side quality enhancement before prediction. We propose a multi-scale dense residual network (MSDRN) to implement both EPI super-resolution and quality enhancement. Transfer learning strategies are used to improve the training performance of quality enhancement. The experimental results show the superior performance of the proposed framework over existing methods in terms of rate-distortion performance.INDEX TERMS Light field image, sparse compression, EPI super-resolution, quality enhancement, convolutional neural network.

show abstract

Section: Effectiveness Of the Lfi Compression Frameworkmentioning

confidence: 65%

Section: Decoder-side Sais Quality Enhancement a Problem Analysismentioning

confidence: 99%

Section: Effectiveness Of the Lfi Compression Frameworkmentioning

confidence: 99%

“…Compared with [7], we consider the influence of compression distortions and add a decoder-side quality enhancement network to improve the performance.…”

Section: Effectiveness Of the Lfi Compression Frameworkmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Light Field Image Compression via CNN-Based EPI Super-Resolution and Decoder-Side Quality Enhancement

Zhao

Yang

et al. 2019

IEEE Access

Self Cite

View full text Add to dashboard Cite

show abstract

Image Compression Scheme Based on Histogram Equalization and Convolution Neural Network

Paul

Chandran

2022

Lecture Notes in Networks and Systems

View full text Add to dashboard Cite

Multi-Models Fusion for Light Field Angular Super-Resolution

Cao

Yang

et al. 2021

ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Self Cite

View full text Add to dashboard Cite

Light field (LF) imaging has received increasing attention due to its richer interpretation of the scene. However, an inherent spatial-angular trade-off exists in LF that prevents LF from practical applications. Consequently, how to break such a trade-off has become one of the main challenges in sparsely sampled LF reconstruction. LF super-resolution (SR) can provide an opportunity to solve this issue, but most methods exploit only one form of LF, thereby leading to much loss of information. We believe that different LF forms can compensate each other to obtain higher gains via fusion strategy. In this paper, therefore, we propose a multi-models fusion for LF SR in angular domain. Cascading models which are trained by different LF forms can fully exploit rich LF information. Experimental results demonstrate that our method is effective and achieves a comparable result against state-of-the-art techniques.

show abstract

Light Field Image Sparse Coding via CNN-Based EPI Super-Resolution

Cited by 10 publications

References 11 publications

Light Field Image Compression via CNN-Based EPI Super-Resolution and Decoder-Side Quality Enhancement

Light Field Image Compression via CNN-Based EPI Super-Resolution and Decoder-Side Quality Enhancement

Image Compression Scheme Based on Histogram Equalization and Convolution Neural Network

Multi-Models Fusion for Light Field Angular Super-Resolution

Contact Info

Product

Resources

About