Binemispheric neuromorphic network

• 

  Binemispheric neuromrphin network

• 

  Test

• 

  High-precision medical image processing. Treatment prediction

• 

  Real-time routing in autonomous vehicles and predictive maintenance

• 

  PCT/IR2025/050026

• 

  Flowchart

Inspiration

The concept of artificial intelligence was first announced in the United States in 1956. When engineers were looking to create computers that could mimic human thought processes, Jack McCarthy claimed that machines could behave like humans. And artificial intelligence means that a computer learns to think like humans and do things, for example: look at images and recognize. The similarity of artificial intelligence and the human brain indicates a search for a better understanding of the functioning of the mind by building systems using information similar to the human brain. Experts have always tried to align this function. But the functioning of the brain has not been accurately simulated. We were inspired by the basic principles and function of the human brain in data processing in the thalamus and the two cerebral hemispheres and the role of the corpus callosum as an interface between the two hemispheres in exchange and coordination, and the initial implementations and results have been very impressive, in fact, separate, simultaneous, coordinated processing and integrated output and the key role in our invention is the information exchange layer between two separate blocks, which in the first stage is divided into two blocks and in subsequent implementations into 8 separate blocks. Because they show technical and operational superiority and efficiency compared to today's models. In addition, in the vision of the system, which is inspired by the neuroscience of the human brain in dividing each hemisphere into 4 separate lobes and the occipital lobe plays a key role in vision, and allows the eagle bird to see objects from long distances with high vision. This feature is due to the special function of the eagle eye, which includes an extended retina, a concave-convex orbit that acts as a lens and enlarges the image and increases visual acuity, so that the reason for this structure and about one and a half million optic nerves. Inspiration from the processing in the human brain and the division of processing and the corpus callosum of the brain played a key role in this invention, which made this network superior in the whole world.

What it does:

In general, this network is used with input text, sound, image, video, classification problems with non-visual data (natural language processing), speech recognition and time series analysis and no restrictions are applied to the input. And it also accepts multi-dimensional data. And it has applications in industry, robotics, autonomous vehicles, medicine, financial markets, agriculture, environment and other fields. Which will be briefly introduced.

1) Smart communication networks: Systems that adjust their performance and optimize communications based on environmental changes. 2) Security systems: Circuits that can encrypt or transmit information based on environmental changes. 3) Advanced applications of this system can be used for electronic skin, advanced sensors or self-organizing documents. 4) Energy management: The use of piezoelectric materials allows the exploitation of environmental energy for the operation of the system.

Electronic skin: Acceptable materials that are controlled by artificial intelligence and can act as skin sensors. 5) Technical and engineering sciences: Predicting electrical load consumption 6) Troubleshooting industrial and technical systems 7) Designing various control systems 8) Designing and optimizing technical and engineering systems 9) Optimal decision-making in engineering projects 10) Financial markets 11) Forecasting stock prices and stock market indices 12) Analysis, evaluation and interpretation of capital and credit 13) Experimental and biological sciences: 14) Predictive modeling of biological and environmental phenomena 15) Identifying hidden and recurring patterns in nature 16) Medical sciences: 17) Modeling of biomedical processes 18) Diagnosing diseases based on the results of medical tests and medical images 19) Predicting treatment outcomes 20) Detecting damage or cracks in engineering structures

How to build it

Why is the invention of the bihemispheric neural network the most fundamental and undeniable invention in the world? One of the most interesting topics that made me think about the architecture of the bihemispheric neural network.

The task was to design and implement a humanoid robot. I was thinking of implementing a robot that behaves like a human. The first possible task was to investigate the existing problems in the field of artificial intelligence and robotic neural networks. Inefficient processing of complex patterns, high energy consumption, high computational cost, low processing speed, interpretability, training, vision, lack of better understanding of patterns and illusions of artificial intelligence were the main problems that I was involved in in this field. With the solutions used in convolutional networks, Möbius transformations in Poincaré space were real edge problems. With the solutions used in convolutional networks, Mobius transformations in Poincaré space, edge problems, information loss, the need for more data and the use of integration and extraction methods, it did not seem like a real topic. In fact, it was a big challenge. Until I turned to studying the biological solution of the initial model of this knowledge. The basic basis that McCarthy, the father of artificial intelligence in America, had first stated. Machines can behave like humans and learn artificial intelligence, that is, computers. Think like humans and do things. Look at images and recognize like humans.

In fact, modern science has always been trying to have brain-like functions. And millions of dollars have been spent on this. But there was no significant result. Looking at brain function, data is amplified, filtered, and categorized in the thalamus of the brain and operates in two output vectors and is processed separately and simultaneously in the two hemispheres with the corpus callosum as the interface and center of information exchange between the two hemispheres. In fact, the duality of the brain's processing, Hebb's law, synaptic plasticity training, the law of two-way communication in the brain, and the dual processing were part of the initial idea. Preprocessing with attention and modulation → two separate output vectors → separate processing → dynamic information exchange → and attention to the outputs of the processing blocks and output integration, proposing the use of attention/modulation/fusion mechanisms or nanoscale transistors that use carbon nanotubes to make artificial intelligence chips.

Simultaneous processing means that In fact, when a person reads a text, the left hemisphere processes words and sentences, and the right hemisphere helps to understand the overall meaning of the text and its connection with the person's previous knowledge. This is an undeniable fact in brain processing.

And in fact, in today's knowledge, the role of the thalamus and how the human brain processes patterns separately and in unison, and the role of the corpus callosum has been ignored. By comparing the technical problems in existing networks and the human brain, the issue becomes quite clear.

Scientific studies in neuroscience show that disorders in the thalamus and the absence of the corpus callosum cause the following to become clear:

1) Expressing false information but believing it (delusions in artificial intelligence)

2) Problem solving problems (weakness in multi-step reasoning and generalization)

3) Problems with advanced activities

4) Difficulty understanding abstract concepts

5) Learning delays

6) Delays in speech and language skills

7) Inability to maintain concentration

8) Difficulty understanding the point of view of others

9) Visual impairments

Slowness in focusing and recognizing complex patterns is mild mental retardation. On this basis, the new network eliminates the illusion of artificial intelligence in the two-hemisphere network in three basic solutions and. On this basis, the two-hemisphere neuromorphic neural network was invented.

Solving technical problems in the field of artificial intelligence and international patent PCT/IR2025/050026 and fundamental and scientific innovation in the new model of the two-hemispheric neural network, piezoelectric memory sensors and smart circuits and solving more problems in the illusion of artificial intelligence, real-time decision-making in robotics and other innovations that were fundamental rather than improving a method, and donating 80% of the benefits of this project and orphan support associations around the world, including me.

Challenges we faced.

Since this project was designed to design a new two-hemisphere network with a preprocessing layer and hidden layers as separate blocks, where data processing is done separately, simultaneously, synchronously, and with integrated output, and the materials presented in other cases were based on patent documents and extensive searches in this regard.

Built With

• payton