Boosting neural video codecs by exploiting hierarchical redundancy

被引:3
|
作者
Pourreza, Reza [1 ]
Le, Hoang [1 ]
Said, Amir [1 ]
Sautiere, Guillaume [1 ]
Wiggers, Auke [1 ]
机构
[1] Qualcomm AI Res, San Diego, CA 92121 USA
关键词
D O I
10.1109/WACV56688.2023.00532
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
In video compression, coding efficiency is improved by reusing pixels from previously decoded frames via motion and residual compensation. We define two levels of hierarchical redundancy in video frames: 1) first-order: redundancy in pixel space, i.e., similarities in pixel values across neighboring frames, which is effectively captured using motion and residual compensation, 2) second-order: redundancy in motion and residual maps due to smooth motion in natural videos. While most of the existing neural video coding literature addresses first-order redundancy, we tackle the problem of capturing second-order redundancy in neural video codecs via predictors. We introduce generic motion and residual predictors that learn to extrapolate from previously decoded data. These predictors are lightweight, and can be employed with most neural video codecs in order to improve their rate-distortion performance. Moreover, while RGB is the dominant colorspace in neural video coding literature, we introduce general modifications for neural video codecs to embrace the YUV420 colorspace and report YUV420 results. Our experiments show that using our predictors with a well-known neural video codec leads to 38% and 34% bitrate savings in RGB and YUV420 colorspaces measured on the UVG dataset.
引用
收藏
页码:5344 / 5353
页数:10
相关论文
共 50 条
  • [41] Alternate video codecs: What are they, and what are they good for?
    Ozur, J
    CD-ROM PROFESSIONAL, 1996, 9 (02): : 82 - &
  • [42] Performance evaluation of video codecs in the space environment
    Tsao, Philip
    Okino, Clayton
    Clare, Loren P.
    2007 IEEE AEROSPACE CONFERENCE, VOLS 1-9, 2007, : 1152 - 1160
  • [43] A comparative analysis of video codecs for multihop wireless video sensor networks
    Imran, Noreen
    Seet, Boon-Chong
    Fong, Alvis C. M.
    MULTIMEDIA SYSTEMS, 2012, 18 (05) : 373 - 389
  • [44] Pattern discrimination method with a boosting approach using hierarchical neural trees
    Okamoto, M.
    Shima, K.
    Matsubara, Y.
    Tsuji, T.
    PROCEEDINGS OF THE INSTITUTION OF MECHANICAL ENGINEERS PART I-JOURNAL OF SYSTEMS AND CONTROL ENGINEERING, 2008, 222 (I7) : 701 - 710
  • [45] Hierarchical video motion estimation using a neural network
    Skrzypkowiak, SS
    Jain, VK
    SECOND INTERNATIONAL WORKSHOP ON DIGITAL AND COMPUTATIONAL VIDEO, PROCEEDINGS, 2001, : 202 - 208
  • [46] Multiple Description Video Coding Based on Hierarchical B Pictures Using Unequal Redundancy
    Tsai, Wen-Jiin
    You, Hao-Yu
    IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY, 2012, 22 (02) : 309 - 320
  • [47] Hierarchical Relation Networks: Exploiting Categorical Structure in Neural Relational Reasoning
    Zou, Ruomu
    Dovrolis, Constantine
    PROCEEDINGS OF THE 13TH INTERNATIONAL JOINT CONFERENCE ON COMPUTATIONAL INTELLIGENCE (IJCCI), 2021, : 359 - 365
  • [48] High Dynamic Range Video Distribution Using Existing Video Codecs
    Touze, David
    Olivier, Yannick
    Thoreau, Dominique
    Serre, Catherine
    2013 PICTURE CODING SYMPOSIUM (PCS), 2013, : 349 - 352
  • [49] A comparative analysis of video codecs for multihop wireless video sensor networks
    Noreen Imran
    Boon-Chong Seet
    Alvis C. M. Fong
    Multimedia Systems, 2012, 18 : 373 - 389
  • [50] EXPLOITING REDUNDANCY TO REDUCE IMPACT FORCE
    KIM, J
    GERTZ, MW
    KHOSLA, PK
    JOURNAL OF INTELLIGENT & ROBOTIC SYSTEMS, 1994, 9 (03) : 273 - 290