interference

Current Source will be ported on .NET Plaform in near future.

Cross-platform C++ library - universal neurobiology-based machine learning framework.

Version 1.1.0

---

PLATFORMS

• Windows (MinGW) (x86)
• Linux (x86, ARM)

---

REQUIREMENTS

• CMake 3.12 or newer
• g++ 7.4.0 or newer (MinGW under Windows)
• LodePNG library (for building inn_vision example only)

---

LICENCE

Interference library is distributed under the MIT Licence.

inn_vision example uses the part of COIL-100 dataset.

"Columbia Object Image Library (COIL-100)," S. A. Nene, S. K. Nayar and H. Murase, Technical Report CUCS-006-96, February 1996. http://www1.cs.columbia.edu/CAVE/software/softlib/coil-100.php

---

HOW TO BUILD

Building library for Windows (MinGW)

Run CMD and follow this steps:

1. Prepare build directory (%INTERFERENCE_ROOT% is root directory of Interference library files, %BUILD_TYPE% is "debug" or "release")

> cd %INTERFERENCE_ROOT%
> mkdir cmake-build-%BUILD_TYPE%
> cd cmake-build-%BUILD_TYPE%

2. Configure build files

> cmake -DCMAKE_BUILD_TYPE=%BUILD_TYPE% -DCMAKE_MAKE_PROGRAM=make.exe -G "MinGW Makefiles" ..

3. Build the library

> make

4. Install the library in comfortable location

> make DESTDIR=.. install

Building library for Linux

Run terminal and follow this steps:

1. Prepare build directory (%INTERFERENCE_ROOT% is root directory of Interference library files, %BUILD_TYPE% is "debug" or "release")

> cd %INTERFERENCE_ROOT%
> mkdir cmake-build-%BUILD_TYPE%
> cd cmake-build-%BUILD_TYPE%

2. Configure build files

> cmake -DCMAKE_BUILD_TYPE=%BUILD_TYPE% ..

3. Build the library

> make

4. Install the library in a global location

> sudo make install

Building an example

To compile an example you need to run this commands:

> cd %INTERFERENCE_ROOT%/samples/%SAMPLE_NAME%
> mkdir cmake-build-%BUILD_TYPE%
> cd cmake-build-%BUILD_TYPE%

Repeat steps 2 and 3 from library building instruction and run make DESTDIR=.. install (for both platforms). Now you can run example using this commands:

> cd ../bin/
> ./%SAMPLE_NAME%_sample

---

RELEASE NOTES

1.1.0

• Interference library performance increased about 15 times
• Added multithreading support for multicore CPUs
• Added experimental support of ARM architecture
• Added ultimate example of image recognition system
• Exceptions raised by inn::Error class are more informative now
• Added doCheckSignal(), doResetSignalController(), getSignal() and getTime() methods for inn::NeuralNet::Link class
• Fixed doAddNeuron(...) method in inn::NeuralNet class (now allowed feedback links)
• Added doComparePatterns() and getNeuronCount() methods for inn::NeuralNet class
• Added new modes of neuron link definition: LinkDefinitionRange and LinkDefinitionRangeNext
• Now getEntries() and getReceptors() methods in inn::Neuron are named as getEntry(...) and getReceptor(...)
• Added doPrepare(), setk1(...) and setk2(...) methods for inn::Neuron, inn::Neuron::Entry and inn::Neuron::Synaps classes
• Added setNeurotransmitterType(...) method for inn::Neuron::Entry and inn::Neuron::Synaps classes
• Added getQSize() method for inn::Neuron::Synaps class
• Added setk3(...) and getdFi() methods for inn::Neuron::Receptor class
• Added getReceptorInfluenceValue(...) and getSynapticSensitivityValue(...) static methods for inn::Neuron::System class
• Added new arithmetic operations methods for inn::Position class
• Added doZeroPosition() and getDistanceFrom(...) methods for inn::Position class

Release notes for all releases can be found in RELEASES.md file.