The paper proposes a novel low-complexity Convolutional Neural Network (CNN) architecture for block-wise angular intra-prediction in lossless video coding. The proposed CNN architecture is designed based on an efficient patch processing layer structure. The proposed CNN-based prediction method is employed to process an input patch containing the causal neighborhood of the current block in order to directly generate the predicted block. The trained models are integrated in the HEVC video coding standard to perform CNN-based angular intra-prediction and to compete with the conventional HEVC prediction. The proposed CNN architecture contains a reduced number of parameters equivalent to only 37% of that of the state-of-the-art reference CNN architecture. Experimental results show that the inference runtime is also reduced by around 5.5% compared to that of the reference method. At the same time, the proposed coding systems yield 83% to 91% of the compression performance of the reference method. The results demonstrate the potential of structural and complexity optimizations in CNN-based intra-prediction for lossless HEVC.
Huang, H, Schiopu, I & Munteanu, A 2020, Low-Complexity Angular Intra-Prediction Con-volutional Neural Network for Lossless HEVC. in IEEE International Workshop on Multimedia Signal Processing. 22 edn, 9287067, IEEE 22nd International Workshop on Multimedia Signal Processing, MMSP 2020, Tampere, pp. 1-6, IEEE International Workshop on Multimedia Signal Processing, Tampere, Finland, 21/09/20. https://doi.org/10.1109/MMSP48831.2020.9287067
Huang, H., Schiopu, I., & Munteanu, A. (2020). Low-Complexity Angular Intra-Prediction Con-volutional Neural Network for Lossless HEVC. In IEEE International Workshop on Multimedia Signal Processing (22 ed., pp. 1-6). Article 9287067 (IEEE 22nd International Workshop on Multimedia Signal Processing, MMSP 2020).. https://doi.org/10.1109/MMSP48831.2020.9287067
@inproceedings{33119caab63e4181b32f9f4ac8426d0e,
title = "Low-Complexity Angular Intra-Prediction Con-volutional Neural Network for Lossless HEVC",
abstract = "The paper proposes a novel low-complexity Convolutional Neural Network (CNN) architecture for block-wise angular intra-prediction in lossless video coding. The proposed CNN architecture is designed based on an efficient patch processing layer structure. The proposed CNN-based prediction method is employed to process an input patch containing the causal neighborhood of the current block in order to directly generate the predicted block. The trained models are integrated in the HEVC video coding standard to perform CNN-based angular intra-prediction and to compete with the conventional HEVC prediction. The proposed CNN architecture contains a reduced number of parameters equivalent to only 37% of that of the state-of-the-art reference CNN architecture. Experimental results show that the inference runtime is also reduced by around 5.5% compared to that of the reference method. At the same time, the proposed coding systems yield 83% to 91% of the compression performance of the reference method. The results demonstrate the potential of structural and complexity optimizations in CNN-based intra-prediction for lossless HEVC.",
keywords = "Deep-learning, low-complexity, angular intra-prediction, lossless video coding",
author = "Hongyue Huang and Ionut Schiopu and Adrian Munteanu",
note = "Funding Information: This research work is funded by Agentschap Innoveren & Ondernemen (VLAIO) within icon project ILLUMINATE HBC.2018.0201 and by Fonds Wetenschappelijk Onderzoek (FWO) - Vlaanderen within project G084117. Publisher Copyright: { extcopyright} 2020 IEEE. Copyright: Copyright 2021 Elsevier B.V., All rights reserved.; IEEE International Workshop on Multimedia Signal Processing, MMSP 2020 ; Conference date: 21-09-2020 Through 24-09-2020",
year = "2020",
month = sep,
day = "21",
doi = "10.1109/MMSP48831.2020.9287067",
language = "English",
series = "IEEE 22nd International Workshop on Multimedia Signal Processing, MMSP 2020",
pages = "1--6",
booktitle = "IEEE International Workshop on Multimedia Signal Processing",
edition = "22",
url = "https://attend.ieee.org/mmsp-2020/",
}