<< Chapter < Page Chapter >> Page >

Universal lossy compression

Consider x α n . The goal of lossy compression  [link] is to describe x ˆ , also of length n but possibly defined over another reconstruction alphabet α ˆ α , such that the description requires few bits and the distortion

d ¯ ( x , x ˆ ) = 1 n i = 1 n d ( x i , x ˆ i )

is small, where d ( · , · ) is some distortion metric. It is well known that for every d ( · , · ) and distortion level D there is a minimum rate R ( D ) , such that x ˆ can be described at rate R ( D ) . The rate R ( D ) is known as the rate distortion (RD) function, it is the fundamental information theoretic limit of lossycompression  [link] , [link] .

The invention of lossy compression algorithms has been a challenging problem for decades. Despite numerous applicationssuch as image compression  [link] , [link] , video compression  [link] , and speech coding  [link] , [link] , [link] , there is a significant gap between theory and practice, and these practical lossy compressorsdo not achieve the RD function. On the other hand, theoretical constructions that achieve the RD function are impractical.

A promising recent algorithm by Jalali and Weissman  [link] is universal in the limit of infinite runtime. Its RD performance is reasonable even with modest runtime.The main idea is that the distortion version x ˆ of the input x can be computed as follows,

x ˆ = arg min w n α n { H k ( w n ) - β d ¯ ( x n , w n ) } ,

where β < 0 is the slope at the particular point of interest in the RD function, and H k ( w n ) is the empirical conditional entropy of order k ,

H k ( w n ) - 1 n a , u k n w ( u k , a ) log n w ( u k , a ) a ' α n w ( u k , a ' ) ,

where u k is a context of order k , and as before n w ( u k , a ) is the number of times that the symbol a appears following a context u k in w n . Jalali and Weissman proved  [link] that when k = o ( log ( n ) ) , the RD pair ( H k ( x ˆ n ) , d ¯ ( x n , x ˆ n ) ) converges to the RD function asymptotically in n . Therefore, an excellent lossy compression technique is to compute x ˆ and then compress it. Moreover, this compression can be universal. In particular, the choice of context order k = o ( log ( n ) ) ensures that universal compressors for context tress sources can emulate the coding length of the empirical conditional entropy H ˆ k ( x ˆ n ) .

Despite this excellent potential performance, there is still a tremendous challenge. Brute force computation of the globally minimum energysolution x n ˆ involves an exhaustive search over exponentially many sequences and is thus infeasible.Therefore, Jalali and Weissman rely on Markov chain Monte Carlo (MCMC)  [link] , which is a stochastic relaxation approach to optimization. The crux of the matter is to definean energy function,

ϵ ( w n ) = H k ( w n ) - β d ( x n , w n ) .

The Boltzmann probability mass function (pmf) is

f s ( w n ) 1 Z t exp { - 1 t ε ( w n ) } ,

where t > 0 is related to temperature in simulated annealing, and Z t is the normalization constant, which does not need to be computed.

Because it is difficult to sample from the Boltzmann pmf [link] directly, we instead use a Gibbs sampler , which computes the marginal distributions at all n locations conditioned on the rest of w n being kept fixed. For each location, the Gibbs sampler resamples from the distribution of w i conditioned on w n i { w n : n i } as induced by the joint pmf in [link] , which is computed as follows,

Questions & Answers

microeconomics deals with individuals and firms while macroeconomics deal with the activities of the society in large
C-Stixxs Reply
Morning guys I need questions under any topic in economics
what is your question
why Economics is science
Fixed Reply
what is economic
Arome Reply
is a science which studies human behavior as a relationship with ends n scarce which have alternative uses
Explain four reason why scale of preference is necessery
Afusat Reply
What's a perfectly elastic demand
jessica Reply
what is the perfect elasticity of demand
what is a balance budget
Azayi Reply
balance budget is when the expenses ND tax re equal
it is when expenditure and revenue are equal
types of cost curves
Lonwabo Reply
there are different types Average total cost curve (ATC) Average Variable cost curve (AVC) Marginal cost curve (MC) .......... and so on
what is histogram
Stella Reply
What is the reason for demand
Taye Reply
hello good afternoon
sheriffvandy Reply
what is production possibility curve
Amina Reply
A curve that shows different combinations (amount) of two goods with a production of limited resources (it is based on the Scarcity principle)
How demand and supply are related
Alimamy Reply
The law of demand states that "all other being equal ",the higher the price the lower the commodity and the lower the price the higher the commodity of purchase within a given period of time .
Emmanuel Reply
participants in market
do you have questions of cerculer flow
do the negative slope of production possibility curve sloped to show trade offs in production of two goods?
Bafokeng Reply
l need the law of demand
The law of demand stated that all other things been equal the high the price the high the quantity demanded and vase visa
The Law of Demand states that ceteris Parabus" The higher the price of the commodity the lower the quantity demanded and the lower the price of the commodity the higher the quantity demanded.
A soccer field is a rectangle 130 meters wide and 110 meters long. The coach asks players to run from one corner to the other corner diagonally across. What is that distance, to the nearest tenths place.
Kimberly Reply
Jeannette has $5 and $10 bills in her wallet. The number of fives is three more than six times the number of tens. Let t represent the number of tens. Write an expression for the number of fives.
August Reply
What is the expressiin for seven less than four times the number of nickels
Leonardo Reply
How do i figure this problem out.
how do you translate this in Algebraic Expressions
linda Reply
why surface tension is zero at critical temperature
I think if critical temperature denote high temperature then a liquid stats boils that time the water stats to evaporate so some moles of h2o to up and due to high temp the bonding break they have low density so it can be a reason
Need to simplify the expresin. 3/7 (x+y)-1/7 (x-1)=
Crystal Reply
. After 3 months on a diet, Lisa had lost 12% of her original weight. She lost 21 pounds. What was Lisa's original weight?
Chris Reply
Difference between extinct and extici spicies
Amanpreet Reply
how did you get the value of 2000N.What calculations are needed to arrive at it
Smarajit Reply
Privacy Information Security Software Version 1.1a
Got questions? Join the online conversation and get instant answers!
Jobilize.com Reply

Get Jobilize Job Search Mobile App in your pocket Now!

Get it on Google Play Download on the App Store Now

Source:  OpenStax, Universal algorithms in signal processing and communications. OpenStax CNX. May 16, 2013 Download for free at http://cnx.org/content/col11524/1.1
Google Play and the Google Play logo are trademarks of Google Inc.

Notification Switch

Would you like to follow the 'Universal algorithms in signal processing and communications' conversation and receive update notifications?