
Intel® Integrated Performance Primitives for Signal Processing
The first operational domain of Intel IPP is primitive functions for
processing "signals", which is an umbrella term for onedimensional
data. Often Intel IPP for signal processing are referred to as Intel IPPS
functions, by the prefix that the function name carries. Most memory,
utility, arithmetic, and conversion functions fall into this category.
The signals domain can be further subdivided into several functional
groups:
List of functions 
Description

Include functions that initialize the values of the elements
of a vector. All vector's elements can be initialized to
zero or to another specified value. They can also be initialized
to the value of a second vector. Samplegenerating functions,
filed in this same group, generate the tone samples, triangle
samples, pseudorandom samples with uniform distribution,
and pseudorandom samples with Gaussian distribution, and
specific Jaehne test samples.

Perform logical and shift, arithmetic, conversion, companding,
windowing, and statistical operations.

Perform convolution and correlation operations, as well
as linear and nonlinear filtering.

Perform the discrete Fourier transform (DFT), the fast
Fourier transform (FFT), the discrete cosine transform (DCT),
as well as well as Hilbert and wavelet transforms of signals.

Perform a series of operations essential for speech recognition
application. This group includes
 functions for generic arithmetic operations on vectors
and matrices
 feature processing functions that preprocess raw speech
signals
 functions for acoustic and language model evaluation,
estimation, and adaptation
 functions for vector quantization and codebook operations
commonly used in acoustic and language model compressions
 Polyphrase sampling functions
 Advanced Aurora functions
 EphraimMalah noise supressing functions
 Acoustic echo canceler functions
 Voice activity detector functions.

The subset of Intel® IPP for audio coding includes
general purpose functions applicable in several codecs and
a number of specific functions for MPEG4 audio encoder
and decoder, MP3 encoder and decoder. These functions implement
pipeline blocks with large computational complexity.
The current set of functions is sufficient to implement
a portable optimized MPEG4 AAC Main profile decoder and
a portable optimized MPEG1, 2 Layer III encoder.
Also, this chapter includes description of companding functions
that perform the µlaw and Alaw companding in compliance
with the CCITT G.711 specification.

Used for implementing speech codecs which follow
ITUT recommendations G.729, G.723.1, G.722.1, G.726
and G.728, as well as GSMAMR, WB GSMAMR and GSMFR
codecs.
When properly built, such speech codecs can be compliant
with the bitexact specifications for published test vectors.
Classified by major functionality, this group includes:
 Common functions
 G.729 related functions (including floatingpoint
functions)
 G.723.1 related functions
 GSMAMR related functions
 WB GSMAMR (G.722.2) related functions
 GSM full rate related runctions
 G.722.1 related runctions
 G.726 related functions
 G.728 related functions.

Take an input vector as argument, compute values of the
respective elementary function elementwise, and return
the results in an output vector.

Perform operations with strings. Intel IPP string functions
do not consider zero as the end of the string, but require
that the length of the string (number of elements) be specified
explicitly.
Overlapping of the strings is not supported (for not inplace
operations).
Intel IPP string functions operate with two data types,
Ipp8u
and Ipp16u.


Click
one of the function types to the left to see a description
here. 


