Pragma Etimos expands patent on audio technology in USA

AI and Audio Recognition are key parts of Pragma Etimos‘ Core Business.

The potential of audio recognition and AI’s predictive abilities is driving efforts to perfect our tools.

After obtaining the patent on “method of processing an audio stream for the recognition of voices and/or background sounds and related system“, applying neural sciences to speech processing, we decided to expand our intellectual boundaries by requesting in the USA the patent just cited.

In this, and in other insights that we will dedicate shortly to the topic, we will briefly analyse the reasons for this overseas projection.

It is the result of a revolutionary technological invention which we believe lays the foundations for the implementation of an IOT device through the biometric control of a non-clonable voice.

The technology behind the patent

Our patent concerns, preliminarily, the biometric recognition of a human voiceprint, overcoming limits and critical issues that international standards had identified, until now, in the comparison of parametric formants.

The audio stream processing method developed represents an innovative solution for the simultaneous recognition of voices and background sounds, offering a highly efficient and precise system.

The technology uses neuronal models, signal processing algorithms, and machine learning to analyze and distinguish elements in an audio stream.

Among the main components of the system there are:

• Pre-processing of the audio signal, further divided into:
  • Noise canceling filter that uses complex algorithms to eliminate interference and improve the quality of the input audio signal.
  • Audio normalization that equalizes the volume level for subsequent analysis.

• Voice Recognition, which allows you to:

  • extract distinctive characteristics of voices, such as formants, vocal signature and individual speaker features.

  • use neural networks (DNNs) trained on large datasets of voices to identify and separate the main voice from the rest of the audio signal so that recognition operations can be performed.

• Sound analysis, to analyse further characteristics of the audio:
  • Segmentation which divides the audio stream into time segments for more accurate identification of background sounds.
  • Classification which applies classification algorithms to categorize the audio file into different categories (age, gender, language, subdivision of multiple speakers…)

This description aims to simplify the operation of a complex technology which required years of social research, analysis, experimentation and development.

The combination of advanced audio signal processing techniques and Artificial Intelligence algorithms offer a robust and versatile system, capable of tackling the challenges of speech recognition in complex environments.

With this innovation, we aim to improve sound quality and reliability in sectors like tactical and strategic security, investigation, forensic studies, commercial applications, and industrial processes for activating, deactivating, and/or neutralizing IoT devices.

Why the American market?

The US technology market is advanced and dynamic, rich in innovation and investment. It continuously develops new solutions in a complex, consumerist, and liberal context where security and self-defense are key segments, presenting challenges that traditional tools can’t address.

In this social and commercial scenario, the integration of AI and Machine Learning into audio technologies is transforming the way devices interpret and interact with sound. Here applications such as advanced speech recognition, sentiment analysis or user experience personalization are growing rapidly.

Smart devices such as smart speakers (e.g. Amazon Echo, Google Home) and virtual assistants are becoming increasingly popular. These not only play music, but interact with users and control other smart home tools.

Our inventions show significant technical-functional differences in this sector, prioritizing security in all technological processes and avoiding web connections to prevent external data contamination.

Alongside the basic IT architecture, the speaker enabled for IoT dialogue is uniquely identified by their voiceprint. Each audio command interaction can only be activated by the authorized subject(s). This distinction is evident compared to smart speakers.

Additionally, active noise cancellation technologies are becoming standard in high-end products, improving the listening experience in noisy environments.

The audio technology market in the United States is booming, driven by continuous innovations and a growing demand for advanced audio solutions. With the integration of AI, Machine Learning and immersive technologies, we strongly believe that our audio patent can represent a strategic asset.

Most likely, the companies that will be able to exploit these trends and maintain a focus on innovation will be the ones that will lead the market in the coming years.

Pragma Etimos patent expansion

Expanding Pragma Etimos‘ audio technology patent to the United States offers a crucial opportunity for enhanced intellectual property protection, access to a large innovative market, and collaborations with industry leaders to maximize the innovation’s potential.

You may also like