Beyond Discretization: A Continuous-Time Framework for Event Generation in Neuromorphic Pixels

2025/04/03 by Hendrickson, Aaron J., Haefner, David P.
#Applications (stat.AP) #FOS: Computer and information sciences

paper · doi:10.48550/arxiv.2504.02803

Abstract

A novel continuous-time framework is proposed for modeling neuromorphic image sensors in the form of an initial canonical representation with analytical tractability. Exact simulation algorithms are developed in parallel with closed-form expressions that characterize the model's dynamics. This framework enables the generation of synthetic event streams in genuine continuous-time, which combined with the analytical results, reveal the underlying mechanisms driving the oscillatory behavior of event data presented in the literature.

Related

• Dynamic Reconstruction from Neuromorphic Data
• Discrete Event System Modeling of Neuromorphic Circuits
• Neuromorphic Sampling of Signals in Shift-Invariant Spaces
• A Sparse Coding Multi-Scale Precise-Timing Machine Learning Algorithm for Neuromorphic Event-Based Sensors
• Covariant spatio-temporal receptive fields for spiking neural networks