Communications of the ACM
Researchers Develop Brain-Inspired Replay Code for Continual Learning
Accuracy on the team's model was about two times higher than that of traditional continual learning models.
Credit: Getty Images
A team of researchers from the Singapore University of Technology and Design designed code to realize an energy-efficient continual learning system. They describe their work in the journal Advanced Theory and Simulations.
The team proposed Brain-Inspired Replay (BIR), which performs continual learning naturally. The BIR model, based on the use of an artificial neural network and a variational autoencoder, imitates the functions of the human brain and can perform well in class-incremental learning situations without storing data.
Led by principal investigator Desmond Loke, an assistant professor at SUTD, the team included Shao-Xiang Go, Qiang Wang, Bo Wang, Yu Jiang, and Natasa Bajalovic.
"A state-of-the-art accuracy of 89% on challenging compliance to current learning tasks without storing data was achieved, which is about two times higher than that of traditional continual learning models, as well as high energy efficiency," Loke says.
From Singapore University of Technology and Design
View Full Article
No entries found