RESEARCHERS at the Jinhua Advanced Research Institute and Harbin University of Science and Technology in China have developed a deep learning algorithm that can detect depression from a person's voice. The tool was trained to recognize emotions in human speech by analysing different vocal characteristics commonly found in the speech of patients diagnosed with depression.
The research, published in the journal Mobile Network and Applications, appears to be promising. The detection rate was 87% for male patients and 87.5% for female patients. To achieve these results, study participants were asked by a virtual agent about their mood and personal life. From this information, the algorithm identified verbal and non-verbal indicators of depression.
"First, the speech information is preprocessed, including speech signal pre-emphasis, framing windowing, endpoint detection, noise reduction, etc.", the study explains.
Several technologies such as 3D and auditory recognition were used to design this tool. For example, the research team used the DAIC-WOZ dataset, which is a collection of audio recordings and 3D facial expressions of patients diagnosed with depression disorders and people without depression.
Similarly, a toolkit called OpenSmile is used to extract speech features from voice recordings.
In the long term, it is possible that this algorithm could serve as an additional tool and an aid in the diagnosis of depression. It could also encourage other researchers to investigate developing artificial intelligence to detect other psychiatric conditions.
Previous studies had already demonstrated the ability of artificial intelligence to detect depressive disorders. In 2021, a team of researchers from the Stevens Institute of Technology in the United States managed to detect depression from texts posted on social networks using AI. According to the WHO, 5% of adults suffer from depression worldwide. – ETX Daily Up, February 19, 2023
_-_SYEDA_IMRAN_-_26.jpg)
