2016/03/14 by Saeed Afshar, Gregory Cohen, Afshar, Saeed +7
Computer Science · Engineering · Neuroscience · #Advanced Memory and Neural Computing #CCD and CMOS Imaging Sensors #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Neural Networks and Applications #Neural and Evolutionary Computing (cs.NE) #Neural dynamics and brain function #Target Tracking and Data Fusion in Sensor Networks
paper · pdf · doi:10.48550/arxiv.1603.04223
openalex publication_date 2016/03/14 · openalex created_date 2022/10/03 · openalex updated_date 2026/07/28
In this paper we compare event-based decaying and time based-decaying memory\nsurfaces for high-speed eventbased tracking, feature extraction, and object\nclassification using an event-based camera. The high-speed recognition task\ninvolves detecting and classifying model airplanes that are dropped free-hand\nclose to the camera lens so as to generate a challenging dataset exhibiting\nsignificant variance in target velocity. This variance motivated the\ninvestigation of event-based decaying memory surfaces in comparison to\ntime-based decaying memory surfaces to capture the temporal aspect of the\nevent-based data. These surfaces are then used to perform unsupervised feature\nextraction, tracking and recognition. In order to generate the memory surfaces,\nevent binning, linearly decaying kernels, and exponentially decaying kernels\nwere investigated with exponentially decaying kernels found to perform best.\nEvent-based decaying memory surfaces were found to outperform time-based\ndecaying memory surfaces in recognition especially when invariance to target\nvelocity was made a requirement. A range of network and receptive field sizes\nwere investigated. The system achieves 98.75% recognition accuracy within 156\nmilliseconds of an airplane entering the field of view, using only twenty-five\nevent-based feature extracting neurons in series with a linear classifier. By\ncomparing the linear classifier results to an ELM classifier, we find that a\nsmall number of event-based feature extractors can effectively project the\ncomplex spatio-temporal event patterns of the dataset to an almost linearly\nseparable representation in feature space.\n