The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Pengekodan dan penghantaran video digital konvensional tidak cekap untuk muat naik video berbilang paparan sumber ramai disebabkan penggunaan kuasa yang tinggi dan kemerosotan kualiti yang tidak diingini dalam saluran wayarles yang tidak stabil. Skim penghantaran video lembut yang dikenali sebagai SoftCast melangkau pengekodan dan penghantaran video digital untuk mengurangkan penggunaan kuasa dalam pengekodan dan penghantaran video. Di samping itu, ia mencapai peningkatan kualiti yang anggun dengan peningkatan kualiti saluran wayarles dengan menghantar terus isyarat video yang diubah suai linear. Walau bagaimanapun, terdapat dua isu biasa untuk menggunakan penghantaran video lembut konvensional kepada muat naik video berbilang tontonan sumber ramai. Pertama, memandangkan penghantaran video lembut telah direka bentuk untuk laluan terus antara setiap penyumbang dan titik akses (AP), ia mungkin mengalami kualiti video yang rendah apabila penyumbang memuat naik videonya ke AP melalui laluan wayarles terus yang tidak stabil. Kedua, penyampaian video lembut konvensional mungkin mengalami kualiti video yang rendah disebabkan oleh penghantaran berlebihan video berkorelasi kerana ia tidak mengeksploitasi korelasi antara kamera yang wujud dalam video berbilang tontonan. Dalam kertas kerja ini, kami mencadangkan skim muat naik video ubah hala berasaskan kluster untuk penstriman video berbilang tontonan sumber orang ramai yang berkualiti tinggi dan berkuasa rendah. Skim yang dicadangkan itu menyepadukan empat pendekatan pengelompokan rangkaian, pemilihan perwakilan, penghantaran video lembut dan transformasi konsine diskret empat dimensi (4D-DCT) untuk memuat naik secara langsung video yang ditangkap ke AP. Khususnya, pengelompokan rangkaian dan pemilihan perwakilan memanfaatkan laluan ubah hala antara penyumbang dan AP. Penghantaran video lembut menghilangkan pengekodan dan penghantaran digital yang haus kuasa dengan menghantar terus pekali domain frekuensi menggunakan DCT berbilang dimensi dan modulasi hampir analog. 4D-DCT mengeksploitasi korelasi kandungan antara penyumbang untuk mengurangkan penghantaran berlebihan. Keputusan penilaian menunjukkan bahawa skim cadangan kami mengatasi skema penyampaian video lembut konvensional apabila perbezaan kualiti saluran antara laluan langsung dan ubah hala meningkat. Selain itu, skim kami mengatasi skim muat naik video berasaskan digital dari segi kualiti video dan penggunaan kuasa. Sebagai contoh, skim yang dicadangkan menghasilkan peningkatan kualiti yang anggun dengan peningkatan kualiti saluran wayarles, bagaimanapun, skim berasaskan digital mengalami kemerosotan kualiti secara tiba-tiba disebabkan oleh ralat penyegerakan dalam penyahkodan.
Than Than NU
Osaka University
Takuya FUJIHASHI
Osaka University
Takashi WATANABE
Osaka University
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Salinan
Than Than NU, Takuya FUJIHASHI, Takashi WATANABE, "Soft Video Uploading for Low-Power Crowdsourced Multi-view Video Streaming" in IEICE TRANSACTIONS on Communications,
vol. E103-B, no. 5, pp. 524-536, May 2020, doi: 10.1587/transcom.2019EBP3134.
Abstract: The conventional digital video encoding and transmission are inefficient for crowdsourced multi-view video uploading due to its high power consumption, and undesirable quality degradation in unstable wireless channel. Soft video delivery scheme known as SoftCast skips digital video encoding and transmission to decrease power consumption in video encoding and transmission. In addition, it achieves graceful quality improvement with the improvement of wireless channel quality by directly sending linear-transformed video signals. However, there are two typical issues to apply conventional soft video delivery to crowdsourced multi-view video uploading. First, since soft video delivery has been designed for direct path between each contributor and the access point (AP), it may suffer low video quality when a contributor uploads its video to the AP over unstable direct wireless path. Second, conventional soft video delivery may suffer low video quality due to the redundant transmission of correlated videos because it does not exploit inter-camera correlations existed in multi-view videos. In this paper, we propose a cluster-based redirect video uploading scheme for high-quality and low-power crowdsourced multi-view video streaming. The proposed scheme integrates the four approaches of network clustering, delegate selection, soft video delivery, and four-dimensional discrete consine transform (4D-DCT) to redirectly upload the captured videos to the AP. Specifically, network clustering and delegate selection leverage the redirect path between the contributors and the AP. Soft video delivery removes power-hungry digital encoding and transmission by directly sending frequency-domain coefficients using multi-dimensional DCT and near-analog modulation. 4D-DCT exploits the content correlations between the contributors to reduce redundant transmissions. Evaluation results show that our proposed scheme outperforms the conventional soft video delivery scheme when the channel quality difference between the direct and redirect paths increases. In addition, our scheme outperforms the digital-based video uploading schemes in terms of both video quality and power consumption. For example, the proposed scheme yields graceful quality improvement with the improvement of wireless channel quality, however, the digital-based schemes suffer from sudden quality degradation due to synchronization errors in decoding.
URL: https://global.ieice.org/en_transactions/communications/10.1587/transcom.2019EBP3134/_p
Salinan
@ARTICLE{e103-b_5_524,
author={Than Than NU, Takuya FUJIHASHI, Takashi WATANABE, },
journal={IEICE TRANSACTIONS on Communications},
title={Soft Video Uploading for Low-Power Crowdsourced Multi-view Video Streaming},
year={2020},
volume={E103-B},
number={5},
pages={524-536},
abstract={The conventional digital video encoding and transmission are inefficient for crowdsourced multi-view video uploading due to its high power consumption, and undesirable quality degradation in unstable wireless channel. Soft video delivery scheme known as SoftCast skips digital video encoding and transmission to decrease power consumption in video encoding and transmission. In addition, it achieves graceful quality improvement with the improvement of wireless channel quality by directly sending linear-transformed video signals. However, there are two typical issues to apply conventional soft video delivery to crowdsourced multi-view video uploading. First, since soft video delivery has been designed for direct path between each contributor and the access point (AP), it may suffer low video quality when a contributor uploads its video to the AP over unstable direct wireless path. Second, conventional soft video delivery may suffer low video quality due to the redundant transmission of correlated videos because it does not exploit inter-camera correlations existed in multi-view videos. In this paper, we propose a cluster-based redirect video uploading scheme for high-quality and low-power crowdsourced multi-view video streaming. The proposed scheme integrates the four approaches of network clustering, delegate selection, soft video delivery, and four-dimensional discrete consine transform (4D-DCT) to redirectly upload the captured videos to the AP. Specifically, network clustering and delegate selection leverage the redirect path between the contributors and the AP. Soft video delivery removes power-hungry digital encoding and transmission by directly sending frequency-domain coefficients using multi-dimensional DCT and near-analog modulation. 4D-DCT exploits the content correlations between the contributors to reduce redundant transmissions. Evaluation results show that our proposed scheme outperforms the conventional soft video delivery scheme when the channel quality difference between the direct and redirect paths increases. In addition, our scheme outperforms the digital-based video uploading schemes in terms of both video quality and power consumption. For example, the proposed scheme yields graceful quality improvement with the improvement of wireless channel quality, however, the digital-based schemes suffer from sudden quality degradation due to synchronization errors in decoding.},
keywords={},
doi={10.1587/transcom.2019EBP3134},
ISSN={1745-1345},
month={May},}
Salinan
TY - JOUR
TI - Soft Video Uploading for Low-Power Crowdsourced Multi-view Video Streaming
T2 - IEICE TRANSACTIONS on Communications
SP - 524
EP - 536
AU - Than Than NU
AU - Takuya FUJIHASHI
AU - Takashi WATANABE
PY - 2020
DO - 10.1587/transcom.2019EBP3134
JO - IEICE TRANSACTIONS on Communications
SN - 1745-1345
VL - E103-B
IS - 5
JA - IEICE TRANSACTIONS on Communications
Y1 - May 2020
AB - The conventional digital video encoding and transmission are inefficient for crowdsourced multi-view video uploading due to its high power consumption, and undesirable quality degradation in unstable wireless channel. Soft video delivery scheme known as SoftCast skips digital video encoding and transmission to decrease power consumption in video encoding and transmission. In addition, it achieves graceful quality improvement with the improvement of wireless channel quality by directly sending linear-transformed video signals. However, there are two typical issues to apply conventional soft video delivery to crowdsourced multi-view video uploading. First, since soft video delivery has been designed for direct path between each contributor and the access point (AP), it may suffer low video quality when a contributor uploads its video to the AP over unstable direct wireless path. Second, conventional soft video delivery may suffer low video quality due to the redundant transmission of correlated videos because it does not exploit inter-camera correlations existed in multi-view videos. In this paper, we propose a cluster-based redirect video uploading scheme for high-quality and low-power crowdsourced multi-view video streaming. The proposed scheme integrates the four approaches of network clustering, delegate selection, soft video delivery, and four-dimensional discrete consine transform (4D-DCT) to redirectly upload the captured videos to the AP. Specifically, network clustering and delegate selection leverage the redirect path between the contributors and the AP. Soft video delivery removes power-hungry digital encoding and transmission by directly sending frequency-domain coefficients using multi-dimensional DCT and near-analog modulation. 4D-DCT exploits the content correlations between the contributors to reduce redundant transmissions. Evaluation results show that our proposed scheme outperforms the conventional soft video delivery scheme when the channel quality difference between the direct and redirect paths increases. In addition, our scheme outperforms the digital-based video uploading schemes in terms of both video quality and power consumption. For example, the proposed scheme yields graceful quality improvement with the improvement of wireless channel quality, however, the digital-based schemes suffer from sudden quality degradation due to synchronization errors in decoding.
ER -