Scientists from Nanyang Technological University in Singapore have presented a system of artificial
In order to increase the recognition accuracygesture, a team of scientists has created a "bio-induced" data fusion system that uses skin-like stretchable sensors made from single-walled carbon nanotubes. In this case, AI processes data from skin and vision in a single place.
This is how a system appeared that combinesThree neural network approaches: the scientists used a "convolutional neural network", which is a machine learning method for early visual processing, a multilayer neural network for early processing of somatosensory information, and a "sparse neural network" for fusing visual and somatosensory information together.
A surgical material based on eggshells has been created
The result is a system that canrecognize gestures more accurately and efficiently than existing methods. It can detect more than 95% of gestures. The team, consisting of scientists from Singapore and the University of Technology Sydney (UTS), published their results in the scientific journal Nature Electronics.
High accuracy was preserved during testingnew AI system in adverse conditions, including noise and poor lighting. The AI system worked effectively in the dark, achieving recognition accuracy of over 96.7%.
Read also
- After the introduction of the Russian vaccine, 144 side effects were found in volunteers
- Created a way to eradicate parasites by blocking all pathways of their metabolism
- Thanks to the new technology, medicines will not deteriorate during temperature extremes