We are developing the deep learning based video coding systems.

 
Screen Shot 2021-03-11 at 12.35.24.png

Title: "RR-DnCNN v2.0: Enhanced Restoration-Reconstruction Deep Neural Network for Down-Sampling Based Video Coding",

Abstract

 we proposed an end-to-end restoration-reconstruction deep neural network (RR-DnCNN) using the degradation-aware technique, which entirely solves degradation from compression and sub-sampling. Besides, we proved that the compression degradation produced by Random Access configuration is rich enough to cover other degradation types, such as Low Delay P and All Intra, for training. Since the straightforward network RR-DnCNN with many layers as a chain has poor learning capability suffering from the gradient vanishing problem, we redesign the network architecture to let reconstruction leverages the captured features from restoration using up-sampling skip connections. Our novel architecture is called restoration-reconstruction u-shaped deep neural network (RR-DnCNN v2.0). As a result, our RR-DnCNN v2.0 outperforms the previous works and can attain 17.02% BD-rate reduction on UHD resolution for all-intra anchored by the standard H.265/HEVC. 

Published: IEEE Transactions on Image Processing (TIP). DOI:10.1109/TIP.2020.3046872. 2021. (SJR Q1, Impact Factor: 9.34) [paper][code]​

 
Screen Shot 2021-03-11 at 12.41.28.png
Screen Shot 2021-03-11 at 12.41.28.png

Title: Image Compression with Encoder-Decoder Matched Semantic Segmentation

Abstract: In recent years, the layered image compression is demonstrated to be a promising direction, which encodes a compact representation of the input image and apply an up-sampling network to reconstruct the image. To further improve the quality of the reconstructed image, some works transmit the semantic segment together with the compressed image data. Consequently, the compression ratio is also decreased because extra bits are required for transmitting the semantic segment. To solve this problem, we propose a new layered image compression framework with encoder-decoder matched semantic segmentation (EDMS). And then, followed by the semantic segmentation, a special convolution neural network is used to enhance the inaccurate semantic segment. As a result, the accurate semantic segment can be obtained in the decoder without requiring extra bits. The experimental results show that the proposed EDMS framework can get up to 35.31% BD-rate reduction over the HEVC-based (BPG) codec, 5% bitrate and 24% encoding time saving compare to the state-of-the-art semantic-based image codec

Published: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), Seattle, WA, USA, 2020, pp. 619-623, doi: 10.1109/CVPRW50498.2020.00088. [paper] [video]

 

Related papers

  • Man M. Ho, Jinjia Zhou*, Gang He, "RR-DnCNN v2.0: Enhanced Restoration-Reconstruction Deep Neural Network for Down-Sampling Based Video Coding", IEEE Transactions on Image Processing (TIP). DOI:10.1109/TIP.2020.3046872. 2021. (SJR Q1, Impact Factor: 9.34)

  • Jinjia Zhou, Victor Sanchez, Lu Zhang, Jianquan Liu, Jiu Xu, "Special Issue on Deep Learning Technologies for Internet of Video Things", IEEE Access, 2021. 

  • Trinh Man Hoang, Jinjia Zhou, and Yibo Fan, “Image Compression with Encoder-Decoder Matched Semantic Segmentation,” 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), Seattle, WA, USA, 2020, pp. 619-623

  • Man M. Ho, Jinjia Zhou, Gang He, Muchen Li, and Lei Li, "SR-CL-DMC: P-frame coding with Super-Resolution, Color Learning, and Deep Motion Compensation", The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops: 3rd Challenge on Learned Image Compression, Seattle, USA, June 2020.

  • Gange He, Chang Wu, Lei Li, Jinjia Zhou, Xianglin Wang, Yunfei Zheng, Bing Yu, and Weiying Xie, "A Video Compression Framework Using an Overfitted Restoration Neural Network", The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops: 3rd Challenge on Learned Image Compression, Seattle, USA, June 2020.

  • Muchen Li, Jinjia Zhou, Satoshi Goto, “A configurable fixed-complexity IME-FME Cost ratio based Inter mode filtering method in HEVC encoding”, IIEEJ Transactions on Image Electronics and Visual Computing, Vol. 8, No. 1, page58-70 (2020.6).

  • Man M. Ho, Gang He, Zheng Wang, and Jinjia Zhou*, "Down-Sampling Based Video Coding with Degradation-aware Restoration-Reconstruction Deep Neural Network", The 26th International Conference on Multimedia Modeling (MMM), Daejeon, Korea, Jan. 2020.DOI: 10.1007/978-3-030-37731-1_9 (Best Paper Runner-Up Award, Oral Acceptance Rate 23.39%, Top-2 Rate: 1.17%)

  • T. M. Hoang and J. Zhou, “B-DRRN: A Block Information Constrained Deep Recursive Residual Network for Video Compression Artifacts Reduction,” 2019 Picture Coding Symposium (PCS), Ningbo, China, 2019, pp. 1-5

  • Do Kim Chi Pham, Jinjia Zhou*, "Deep Learning-based Luma and Chroma Fractional Interpolation in Video Coding", IEEE Access, Vol. 7, pp. 112535-112543, Aug. 2019.  DOI: 10.1109/ACCESS.2019.2935378 (SJR Q1, Impact Factor:4.098)

  • Pham Do Kim Chi, Jinjia Zhou, “A Convolutional Neural Network for Fractional Interpolation in Video Coding ”, The International Symposium on Artificial Intelligence and Robotics (ISAIR 2019), Daegu, Korea, Aug. 2019. (Best Paper Nomination).