{"id":12138,"date":"2022-08-31T20:30:32","date_gmt":"2022-08-31T20:30:32","guid":{"rendered":"http:\/\/TheNextWeb=1391126"},"modified":"2022-08-31T20:30:32","modified_gmt":"2022-08-31T20:30:32","slug":"researchers-in-italy-and-germany-unveil-neuromorphic-approach-to-robotics","status":"publish","type":"post","link":"https:\/\/www.londonchiropracter.com\/?p=12138","title":{"rendered":"Researchers in Italy and Germany unveil neuromorphic approach to robotics"},"content":{"rendered":"\n

Scientists have tapped neuromorphic computing<\/a> to keep robots<\/a> learning about new objects after they\u2019ve been deployed.<\/span><\/p>\n

For the uninitiated, neuromorphic computing replicates the neural structure of the human brain to create algorithms<\/a> that can deal with the uncertainties of the natural world.<\/p>\n

Intel<\/a> Labs has developed one of the most notable architectures in the field: the Loihi<\/a> neuromorphic chip.<\/span><\/p>\n

\n
\n
<\/div>\n

\n

Greetings, humanoids<\/h2>\n

Subscribe to our newsletter now for a weekly recap of our favorite AI stories in your inbox.<\/p>\n<\/div>\n<\/div>\n<\/div>\n

Loihi is comprised of around 130,000 artificial neurons, which send information to each other across a \u201cspiking\u201d neural network (SNN). The chips had already powered a range of systems, from an smart artificial skin<\/a> to an electronic \u201cnose\u201d<\/a> that recognizes scents emitted from explosives.<\/span><\/p>\n

Intel Labs this week unveiled another application. The research unit teamed up with the Italian Institute of Technology and the Technical University of Munich to deploy Loihi in a new approach to continual learning for robotics.<\/span><\/p>\n

Interactive learning<\/h2>\n

The method targets systems that interact with unconstrained environments, such as future robotic assistants for healthcare and manufacturing.<\/p>\n

Existing deep neural networks can struggle with object learning in these scenarios, as they require extensive well-prepared training data \u2014 and careful retraining on new objects they encounter. The new neuromorphic approach aims to overcomes these limitations.<\/p>\n

The researchers first implemented an SNN on Loihi. This architecture localizes learning to a single layer of plastic synapses. It also accounts for different views of objects by adding new neurons on demand. <\/span><\/p>\n

As a result, the learning process unfolds autonomously while interacting with the user.<\/span><\/p>\n

Neuromorphic simulations<\/h2>\n

The team tested their approach in a simulated 3D environment. In this setup, the robot actively senses objects by moving an event-based camera that functions as its eyes. <\/span><\/p>\n

The camera\u2019s sensor \u201csees\u201d objects in a manner inspired by small fixational eye movements called \u201cmicrosaccades.\u201d  If the object it views is new, the SNN representation is learned or updated. If the object  is known, the network recognizes it and provides feedback to the user.<\/span><\/p>\n

\"In
<\/i><\/a>The data collected by the camera drives an SNN on the Loihi chip. Credit: Intel Labs<\/figcaption>