Neuromorphic (or event-based) sensors capture events at a microsecond resolution, which requires low-latency processing by FPGAs for real-time performance. By using the temporal and spatial components of the events, some algorithms, like the Hierarchy of Event-Based Time-Surfaces, find relationships between the two components for enhanced feature extraction and object detection. In contrast, using K-means clustering on spatial information can explore the tradeoff between increased accuracy at the expense of performance.

Speaker

Luke Kljucaric is the lead computer engineering PhD student of the HPC/Reconfigurable Systems Group in the National Science Foundation (NSF) at the Center for Space, High-Performance, and Resilient Computing (SHREC) with the University of Pittsburgh.

His research focuses on accelerated machine learning, which includes a traditional and neuromorphic classification studied on CPUs, GPUs, and FPGAs.