Learning Patterns of Latent Residual for Improving Video Compression

Yen-Chung Chen; Keng-Jui Chang; Yi-Hsuan Tsai; Wei-Chen Chiu

Learning Patterns of Latent Residual for Improving Video Compression

Yen-Chung Chen, Keng-Jui Chang, Yi-Hsuan Tsai, Wei-Chen Chiu

資訊工程學系

研究成果: Conference contribution › 同行評審

摘要

We tackle the problem of reducing compression artifacts. Specifically, we focus on transmitting the residual from the original video, i.e. difference between a compressed video and its corresponding original/uncompressed one, together with the compressed video during video transmission. Our video compression pipeline is capable of diminishing the overall cost of transmitting the residual and simultaneously achieving comparable video quality with respect to a state-of-the-art baseline. We provide experimental results on several datasets, including the one with great diversity, to substantiate the capacity of our pipeline in improving video compression.

原文	English
主出版物標題	The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops
頁數	5
出版狀態	Published - 六月 2019

指紋深入研究「Learning Patterns of Latent Residual for Improving Video Compression」主題。共同形成了獨特的指紋。

引用此

@inproceedings{e430bd38099247469ea8f9e86649fcde,

title = "Learning Patterns of Latent Residual for Improving Video Compression",

abstract = "We tackle the problem of reducing compression artifacts. Specifically, we focus on transmitting the residual from the original video, i.e. difference between a compressed video and its corresponding original/uncompressed one, together with the compressed video during video transmission. Our video compression pipeline is capable of diminishing the overall cost of transmitting the residual and simultaneously achieving comparable video quality with respect to a state-of-the-art baseline. We provide experimental results on several datasets, including the one with great diversity, to substantiate the capacity of our pipeline in improving video compression.",

author = "Yen-Chung Chen and Keng-Jui Chang and Yi-Hsuan Tsai and Wei-Chen Chiu",

year = "2019",

month = jun,

language = "English",

booktitle = "The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops",

}

TY - GEN

T1 - Learning Patterns of Latent Residual for Improving Video Compression

AU - Chen, Yen-Chung

AU - Chang, Keng-Jui

AU - Tsai, Yi-Hsuan

AU - Chiu, Wei-Chen

PY - 2019/6

Y1 - 2019/6

N2 - We tackle the problem of reducing compression artifacts. Specifically, we focus on transmitting the residual from the original video, i.e. difference between a compressed video and its corresponding original/uncompressed one, together with the compressed video during video transmission. Our video compression pipeline is capable of diminishing the overall cost of transmitting the residual and simultaneously achieving comparable video quality with respect to a state-of-the-art baseline. We provide experimental results on several datasets, including the one with great diversity, to substantiate the capacity of our pipeline in improving video compression.

AB - We tackle the problem of reducing compression artifacts. Specifically, we focus on transmitting the residual from the original video, i.e. difference between a compressed video and its corresponding original/uncompressed one, together with the compressed video during video transmission. Our video compression pipeline is capable of diminishing the overall cost of transmitting the residual and simultaneously achieving comparable video quality with respect to a state-of-the-art baseline. We provide experimental results on several datasets, including the one with great diversity, to substantiate the capacity of our pipeline in improving video compression.

M3 - Conference contribution

BT - The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops

ER -