Event News

Talk by Prof. Jeremy Gibbons from Oxford University: "Asymmetric Numeral Systems"

We are pleased to announce that Prof. Jeremy Gibbons from Oxford University will talk on program calculation about

Asymmetric Numeral Systems

on October 17, 2018. Everyone is welcome to join us.


Oct. 17 (Wed.)




National Institute of Informatics
Room. 1517, 15th Floor, NII


Jeremy Gibbons (Oxford University)


Asymmetric Numeral Systems


Asymmetric Numeral Systems (ANS) are a recent entropy encoding technique, compressing text and achieving almost the information-theoretic limit implied by Shannon's Theorem (the number of bits used to encode a symbol is the logarithm of the symbol entropy - and in particular, may be fractional). ANS has a particularly fast implementation, and it's now used in many real-world encoders developed by companies such as Facebook, Apple, and Google. ANS therefore combines the effectiveness of Arithmetic Coding with the efficiency of Huffman Coding. I will present a calculation of ANS encoding and decoding from an abstract specification. This is work in progress - there are gaps in my calculations, with which I would appreciate help!


Jeremy Gibbons is Professor of Computing in the Department of Computer Science at the University of Oxford. He is currently Director of the Software Engineering Programme, which offers part-time professional Masters' degrees in Software Engineering and in Software and Systems Security. He also lead the Algebra of Programming research group. He is Editor-in-Chief of the Journal of Functional Programming, Past Vice Chair of ACM SIGPLAN, Past Chair of IFIP WG2.1.


Zhenjiang Hu ( hu [at] nii.ac.jp )