The Brain-like Powers of AI: Large Scale Spiking Networks for Local Devices

AI study reveals large neural networks that resemble brains

Researchers Bojian Yun and Sander Bohte of the HBP-partner Dutch National Research Institute for Mathematics and Computer Science CWI demonstrate in a new paper in Nature Machine Intelligence a major step in artificial intelligence, which can be applied to local devices such as smartphones and VR-like applications while maintaining privacy.

The researchers demonstrate how novel learning methods combined with brain-like neuronal structures enable the training of large-scale spiking neural network quickly and efficiently. Wearable AI, speech recognition, and Augmented Reality are all potential applications.

Modern artificial neural networks, which are at the core of the AI revolution, are only loosely based on networks of biological neurons, such as the brain. The brain is a larger network that uses less energy and responds ultra-fast to external stimuli. Spiking neural network are special types that mimic the functioning of biological neurons. The neurons in our nervous system exchange electrical pulses sparingly.