A team of MIT scientists has developed an app which could make it easier to read the conversation emotions. This could help people determine if the speaker is actually excited.
Most anyone must have been in the following situation. You are talking with someone. And they are telling you a sad story. Or, on the contrary, a happy one. But some people may make you wonder. Are they actually happy or sad? Now, a team of scientists may have just offered a solution.
The research team comes from the Massachusetts Institute of Technology or MIT. They are part of the Computer Science and Artificial Intelligence Laboratory or CSAIL. They have been involved in an ongoing study.
This targets the study of emotion detection. The scientists have already released studies on the matter. And they also presented some of their developed devices. For example, last fall, they built a new device. This was able to detect human emotions. How? By using wireless signals.
Now, the team took the device one step further. Earlier this week, they presented a software that can detect conversation emotions. This was incorporated into an app. One targeted specifically for wearables.
It should be able to follow a conversation. And as such, identify the emotions behind the dialog. This conversation emotions tool should work for every part of the speech.
During testing, the researchers incorporated the app into a fitness tracker. It should be able to offer information in real time. The app works as follows. During conversations, it collects speech data. And also physical information. It then analyzes the overall tone.
Based on these elements, it should provide the conversation emotions result. And it can also offer part specific information. It can be used to determine if one was sad or another happy.
This app will track the emotions in five-second intervals. As such, it can also track and point out changes. It can determine if the person had a change of tone.
The analyzed physical change will be targeting the following. They will be monitoring the participant’s skin temperature. And also their heart rate or physical movement. Waving or crossing your arms are considered tell-tale signs.
This new app is AI-powered. The AI or artificial intelligence makes use of neural networks. These can analyze and compare huge amounts of data. And they can also acquire need data. By learning, such technology should also become “smarter”. And also offer better results.
The researchers also carried out some field tests. They asked the participants to wear a Simband. Their conversation emotion app had been previously installed. Overall results reported the following. The neural networks were 83 percent accurate in reporting the conversation’s tone.
In most cases, the AI system associated the emotion as follows. Monotonous vocals interrupted by long pauses were considered sad. Speech patterns that varied were considered happy. These are just a few, base emotions.
Research is hoping to further develop the emotions spectrum. They are hoping to add more complex, nuanced feelings. More details about the current software were offered in a study paper. This was released on February 01 by CSAIL. The paper was titled as follows. “Predicting Latent Narrative Mood using Audio and Physiologic Data”.
Research on the matter will continue. According to one of the study members, such an app could have varied uses. It could help people with anxiety. Or even persons diagnosed with medical conditions such as Asperger’s.
Image Source: Wikimedia
Roxanne Briean
Latest posts by Roxanne Briean (see all)
- Former Virginia Tech Freshman Sentenced to 50 Years in Prison for Stabbing a Girl to Death - June 28, 2018
- San Francisco Woman Threatened to Call Police on Girl Who Sold Ice Water for Disneyland Trip - June 25, 2018
- Epping Woman Sentence to 20 Years of Prison for Mutilating Homeless Woman’s Children - June 12, 2018