Deep audio analysis to improve speech recognition

Learning to understand grounded language, language that occurs in the context of, and refers to, the world at large, is a popular area of ​​research in robotics and informatics. Most of the current work in this area still operates with textual data, and that limits the ability to deploy agents in realistic environments.

A recent research article proposes acquiring grounded language directly from the end user’s speech using a relatively small number of data points rather than relying on intermediate textual representations.

AI can improve speech recognition systems

A detailed analysis of the natural language grounding is provided from raw speech to robotic sensor data of everyday objects using state-of-the-art speech representation models. Analysis of the speech and audio qualities of individual participants shows that learning directly from raw speech improves the performance of stressed-voice users compared to relying on automatic transcriptions.

The study of grounded language, which connects natural language with perceptions, is an important research area for this area. Previous work on grounded language acquisition has focused primarily on textual inputs.

Via this investigation it was possible to show the feasibility of performing language acquisition based on paired visual perceptions and raw speech inputs. This will allow interactions in which language about new tasks and environments is learned from end users, reducing reliance on textual inputs and potentially mitigating the effects of demographic bias found in widely available speech recognition systems.

The research team leveraged recent work on self-monitoring speech representation models to show that learned representations of speech can make language grounding systems more inclusive towards specific groups, while maintaining or even increasing performance. general.

Obviously, a system capable of better recognizing the human voice reduces the margin of error in people’s interaction with systems that are increasingly incorporated into daily life, such as virtual assistants, which, in addition to mobile phones, are present in every more and more devices.

Source link

Lenny Li

I started to play with tech since middle school. Smart phones, laptops and gadgets are all about my life. Besides, I am also a big fan of Star War. May the force be with you!

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button