Large-scale Model of Cortex Successfully Solving Visual Tasks: A New Step Towards Neuromorphic Computing

New large-scale virtual cortex model highly successful at solving visual tasks

Researchers at HBP have developed a large-scale mouse primary visual cortex model that can solve visual tasks with a high degree of robustness. This model is the foundation for a whole new generation of neural networks. These models are versatile and can be processed efficiently, which makes them a good candidate for neuromorphic computing.

Modeling the human brain has a huge impact on artificial intelligence. Since the brain is much more efficient at processing images than artificial networks, researchers can take inspiration from the neuroscience field to create neural network that work similarly to biological networks to save significant energy.

Brain-inspired neural network blueprints will likely have a significant impact on the future of technology by allowing for more energy-efficient processing. Researchers from Graz University of Technology’s Human Brain Project (HBP), Austria, have shown in a new study how a large-scale data-based model is able to accurately and versatilely reproduce the visual processing abilities of the brain. The results of the study were published in Science Advances.