The AI ​​can now detect depression in your voice



[ad_1]

New Delhi: India – the sixth most depressed country in the world – would have about 56 million people suffering from depression and 38 million people with anxiety disorders, according to a recent report from the World Health Organization. health (WHO).

To help quickly identify depression, scientists have now improved a technology that uses artificial intelligence to sift the sound of your voice to determine whether you are depressed or not.

Researchers in computer science at the University of Alberta in Canada have improved the technology for identifying depression by voice signals.

The study, led by Mashrura Tasnim and Professor Eleni Stroulia, builds on previous research that suggests that the tone of our voice contains information about our mood.

Using standard reference datasets, Tasnim and Stroulia have developed a methodology combining multiple machine learning (ML) algorithms to more accurately recognize depression with the help of acoustic cues.

A realistic scenario is to ensure that users use an application that collects voice samples when they speak naturally.

"The app, run on the user's phone, recognizes and tracks mood indicators, such as depression, over time. Just as you have a step counter on your phone, you can have a depression indicator based on your voice when you use the phone, "said Stroulia.

The WHO considers that depression is the main contributor to global disability. It is also the main contributor to deaths by suicide.

The ultimate goal, said the researchers, is to develop meaningful applications from this technology.

Such a tool could be useful for supporting work with caregivers or helping people reflect on their own mood over time.

"This work, which involves developing more accurate detection in standard reference datasets, is the first step," added Stroulia at the recent presentation of the document at the Canadian Conference on Artificial Intelligence.

(This story was published from a news agency thread without text modification.Only the title has been changed.)

[ad_2]

Source link