Online Few-shot Gesture Learning on a Neuromorphic Processor
This addresses the challenge of real-time adaptation for neuromorphic computing in applications like gesture recognition, though it is incremental as it builds on existing transfer learning and computational neuroscience principles.
The paper tackles the problem of enabling neuromorphic processors to learn new classes of data quickly with minimal examples, and the result is that the SOEL system achieves online few-shot learning of gestures, including from live sensor data, with faster learning and fewer updates.
We present the Surrogate-gradient Online Error-triggered Learning (SOEL) system for online few-shot learning on neuromorphic processors. The SOEL learning system uses a combination of transfer learning and principles of computational neuroscience and deep learning. We show that partially trained deep Spiking Neural Networks (SNNs) implemented on neuromorphic hardware can rapidly adapt online to new classes of data within a domain. SOEL updates trigger when an error occurs, enabling faster learning with fewer updates. Using gesture recognition as a case study, we show SOEL can be used for online few-shot learning of new classes of pre-recorded gesture data and rapid online learning of new gestures from data streamed live from a Dynamic Active-pixel Vision Sensor to an Intel Loihi neuromorphic research processor.