RNN-based bitstream feature extraction method for codec classification

Seungwoo Wee, Jechang Jeong

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In this paper, we propose codec classification algorithm based on recurrent neural network (RNN) model. In video compression, codecs, such as MPEG2 and H.264/AVC, have their own distinctive data structure. These unique structures which are almost shown in header can be considered their feature. The proposed algorithm exploits that characteristics for classifying unknown bitstreams into specific codec. According to the fact that RNN is appropriate to time series data for learning to classification/recognition, the feature of an encoded bitstream can be extracted. We constitute the encoded bitstream as an input and give the bitstream its label indicating codec index. Two standard codecs, MPEG2 and H.264/AVC, are used in experiment. Experimental results show that the proposed RNN model classified bitstreams into corresponding codecs to some extent.

Original languageEnglish
Title of host publicationInternational Workshop on Advanced Image Technology, IWAIT 2019
EditorsWen-Nung Lie, Kazuya Hayase, Qian Kemao, Lu Yu, Phooi Yee Lau, Sanun Srisuk, Yung-Lyul Lee
PublisherSPIE
ISBN (Electronic)9781510627734
DOIs
StatePublished - 2019 Jan 1
EventInternational Workshop on Advanced Image Technology 2019, IWAIT 2019 - Singapore, Singapore
Duration: 2019 Jan 62019 Jan 9

Publication series

NameProceedings of SPIE - The International Society for Optical Engineering
Volume11049
ISSN (Print)0277-786X
ISSN (Electronic)1996-756X

Conference

ConferenceInternational Workshop on Advanced Image Technology 2019, IWAIT 2019
CountrySingapore
CitySingapore
Period19/01/619/01/9

Fingerprint

Recurrent neural networks
Recurrent Neural Networks
pattern recognition
Feature Extraction
MPEG-2
Feature extraction
Neural Network Model
headers
video compression
Video Compression
data structures
Classification Algorithm
Image compression
Time Series Data
classifying
learning
Data structures
Labels
Time series
Data Structures

Keywords

  • Classification
  • bitstream feature extraction
  • recurrent neural network

Cite this

Wee, S., & Jeong, J. (2019). RNN-based bitstream feature extraction method for codec classification. In W-N. Lie, K. Hayase, Q. Kemao, L. Yu, P. Y. Lau, S. Srisuk, & Y-L. Lee (Eds.), International Workshop on Advanced Image Technology, IWAIT 2019 [110493N] (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 11049). SPIE. https://doi.org/10.1117/12.2521425
Wee, Seungwoo ; Jeong, Jechang. / RNN-based bitstream feature extraction method for codec classification. International Workshop on Advanced Image Technology, IWAIT 2019. editor / Wen-Nung Lie ; Kazuya Hayase ; Qian Kemao ; Lu Yu ; Phooi Yee Lau ; Sanun Srisuk ; Yung-Lyul Lee. SPIE, 2019. (Proceedings of SPIE - The International Society for Optical Engineering).
@inproceedings{ad26bc918482481092c233185c695cbd,
title = "RNN-based bitstream feature extraction method for codec classification",
abstract = "In this paper, we propose codec classification algorithm based on recurrent neural network (RNN) model. In video compression, codecs, such as MPEG2 and H.264/AVC, have their own distinctive data structure. These unique structures which are almost shown in header can be considered their feature. The proposed algorithm exploits that characteristics for classifying unknown bitstreams into specific codec. According to the fact that RNN is appropriate to time series data for learning to classification/recognition, the feature of an encoded bitstream can be extracted. We constitute the encoded bitstream as an input and give the bitstream its label indicating codec index. Two standard codecs, MPEG2 and H.264/AVC, are used in experiment. Experimental results show that the proposed RNN model classified bitstreams into corresponding codecs to some extent.",
keywords = "Classification, bitstream feature extraction, recurrent neural network",
author = "Seungwoo Wee and Jechang Jeong",
year = "2019",
month = "1",
day = "1",
doi = "10.1117/12.2521425",
language = "English",
series = "Proceedings of SPIE - The International Society for Optical Engineering",
publisher = "SPIE",
editor = "Wen-Nung Lie and Kazuya Hayase and Qian Kemao and Lu Yu and Lau, {Phooi Yee} and Sanun Srisuk and Yung-Lyul Lee",
booktitle = "International Workshop on Advanced Image Technology, IWAIT 2019",

}

Wee, S & Jeong, J 2019, RNN-based bitstream feature extraction method for codec classification. in W-N Lie, K Hayase, Q Kemao, L Yu, PY Lau, S Srisuk & Y-L Lee (eds), International Workshop on Advanced Image Technology, IWAIT 2019., 110493N, Proceedings of SPIE - The International Society for Optical Engineering, vol. 11049, SPIE, International Workshop on Advanced Image Technology 2019, IWAIT 2019, Singapore, Singapore, 19/01/6. https://doi.org/10.1117/12.2521425

RNN-based bitstream feature extraction method for codec classification. / Wee, Seungwoo; Jeong, Jechang.

International Workshop on Advanced Image Technology, IWAIT 2019. ed. / Wen-Nung Lie; Kazuya Hayase; Qian Kemao; Lu Yu; Phooi Yee Lau; Sanun Srisuk; Yung-Lyul Lee. SPIE, 2019. 110493N (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 11049).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - RNN-based bitstream feature extraction method for codec classification

AU - Wee, Seungwoo

AU - Jeong, Jechang

PY - 2019/1/1

Y1 - 2019/1/1

N2 - In this paper, we propose codec classification algorithm based on recurrent neural network (RNN) model. In video compression, codecs, such as MPEG2 and H.264/AVC, have their own distinctive data structure. These unique structures which are almost shown in header can be considered their feature. The proposed algorithm exploits that characteristics for classifying unknown bitstreams into specific codec. According to the fact that RNN is appropriate to time series data for learning to classification/recognition, the feature of an encoded bitstream can be extracted. We constitute the encoded bitstream as an input and give the bitstream its label indicating codec index. Two standard codecs, MPEG2 and H.264/AVC, are used in experiment. Experimental results show that the proposed RNN model classified bitstreams into corresponding codecs to some extent.

AB - In this paper, we propose codec classification algorithm based on recurrent neural network (RNN) model. In video compression, codecs, such as MPEG2 and H.264/AVC, have their own distinctive data structure. These unique structures which are almost shown in header can be considered their feature. The proposed algorithm exploits that characteristics for classifying unknown bitstreams into specific codec. According to the fact that RNN is appropriate to time series data for learning to classification/recognition, the feature of an encoded bitstream can be extracted. We constitute the encoded bitstream as an input and give the bitstream its label indicating codec index. Two standard codecs, MPEG2 and H.264/AVC, are used in experiment. Experimental results show that the proposed RNN model classified bitstreams into corresponding codecs to some extent.

KW - Classification

KW - bitstream feature extraction

KW - recurrent neural network

UR - http://www.scopus.com/inward/record.url?scp=85063880593&partnerID=8YFLogxK

U2 - 10.1117/12.2521425

DO - 10.1117/12.2521425

M3 - Conference contribution

AN - SCOPUS:85063880593

T3 - Proceedings of SPIE - The International Society for Optical Engineering

BT - International Workshop on Advanced Image Technology, IWAIT 2019

A2 - Lie, Wen-Nung

A2 - Hayase, Kazuya

A2 - Kemao, Qian

A2 - Yu, Lu

A2 - Lau, Phooi Yee

A2 - Srisuk, Sanun

A2 - Lee, Yung-Lyul

PB - SPIE

ER -

Wee S, Jeong J. RNN-based bitstream feature extraction method for codec classification. In Lie W-N, Hayase K, Kemao Q, Yu L, Lau PY, Srisuk S, Lee Y-L, editors, International Workshop on Advanced Image Technology, IWAIT 2019. SPIE. 2019. 110493N. (Proceedings of SPIE - The International Society for Optical Engineering). https://doi.org/10.1117/12.2521425