Friday, March 8, 2019

Software Based on the Deep Learning System Decoded Rodent Chatter


Image credit: Alice Gray.

DeepSqueak, the program developed by a team of researchers from the University of Washington School of Medicine takes sounds and transforms them into images of the sound, known as sonograms.

The software that was developed by the researchers is the first of a kind artificial neural networks that was used to detect squeaks. The basis of this technology was self-driving cars, and their ability to transform an audio signal as a visual representation.

The squeaks that were produced by rodents depended on the stages of their courtship. For example, when two male mice were put together, they produced the same calls repeatedly. However, when a scent of a female mouse was introduced nearby, the vocalisation got more complex. This suggests that the male mice were singing a courtship song.

This technology would allow us to better understand all the aspects of the change in brain activity when affected by pleasure. In the scale of the study, the pleasure is brought to the mice by anticipation of treats or games with their peers. However, if we apply this concept to the human brain, we can change the treats and games to drugs.

Professor John Neumaier, the senior author of the project, said that their goal is to develop treatments for drugs withdrawal. The researchers hope that with a better understanding of the change in brain activity caused by pressure or troublesome feelings, they would help treat drug addictions. 

No comments:
Write comments