preprint · 2022

Adaptive Context Modeling for Arithmetic Coding Using Perceptron

Lucas Lopes, Philip A. Chou, Ricardo L. de Queiroz

Cited by 0Open access

Abstract

<p>Arithmetic coding is used in most media compression methods. Context modeling is usually done through frequency counting and look-up tables (LUTs). For long-memory signals, probability modeling with large context sizes is often infeasible. Recently, neural networks have been used to model probabilities of large contexts in order to drive arithmetic coders. These neural networks have been trained offline. We introduce an online method for training a perceptron-based context-adaptive arithmetic coder on-the-fly, called adaptive perceptron coding, which continuously learns the context probabilities and quickly converges to the signal statistics.We test adaptive perceptron coding over a binary image database, with results always exceeding the performance of LUT-based methods for large context sizes and of recurrent neural networks. We also compare the method to a version requiring offline training, which leads to equally satisfactory results.</p>

Cite this paper

APA 7

Lopes, L., Chou, P. A., & de Queiroz, R. L. (2022). Adaptive Context Modeling for Arithmetic Coding Using Perceptron. https://doi.org/10.36227/techrxiv.21225629

MLA 9

Lopes, Lucas, et al. “Adaptive Context Modeling for Arithmetic Coding Using Perceptron.” 2022, https://doi.org/10.36227/techrxiv.21225629.

Chicago

Lopes, Lucas, Philip A. Chou, and Ricardo L. de Queiroz. 2022. “Adaptive Context Modeling for Arithmetic Coding Using Perceptron.” https://doi.org/10.36227/techrxiv.21225629.

Harvard

Lopes, L., Chou, P.A. and de Queiroz, R.L. (2022) ‘Adaptive Context Modeling for Arithmetic Coding Using Perceptron’. Available at: https://doi.org/10.36227/techrxiv.21225629.

BibTeX

@misc{lopes2022adaptive,
  title = {{Adaptive Context Modeling for Arithmetic Coding Using Perceptron}},
  author = {Lopes, Lucas and Chou, Philip A. and de Queiroz, Ricardo L.},
  year = {2022},
  doi = {10.36227/techrxiv.21225629}
}

Made from the index record. Check names, capitalisation and page numbers against the paper. Need another style? AI for Research has over 10,000.