Sepia, Tarsier, and Chameleon: A Modular C++ Framework for Event-Based Computer Vision

Alexandre Marcireau; Sio-Hoi Ieng; Ryad Benosman

doi:10.3389/fnins.2019.01338

Sepia, Tarsier, and Chameleon: A Modular C++ Framework for Event-Based Computer Vision

Front Neurosci. 2020 Jan 8:13:1338. doi: 10.3389/fnins.2019.01338. eCollection 2019.

Authors

Alexandre Marcireau¹, Sio-Hoi Ieng¹, Ryad Benosman^{1

2

3}

Affiliations

¹ INSERM UMRI S 968, Sorbonne Universites, UPMC Univ Paris 06, UMR S 968, CNRS, UMR 7210, Institut de la Vision, Paris, France.
² University of Pittsburgh Medical Center, Pittsburgh, PA, United States.
³ Robotics Institute, Carnegie Mellon University, Pittsburgh, PA, United States.

Abstract

This paper introduces an new open-source, header-only and modular C++ framework to facilitate the implementation of event-driven algorithms. The framework relies on three independent components: sepia (file IO), tarsier (algorithms), and chameleon (display). Our benchmarks show that algorithms implemented with tarsier are faster and have a lower latency than identical implementations in other state-of-the-art frameworks, thanks to static polymorphism (compile-time pipeline assembly). The observer pattern used throughout the framework encourages implementations that better reflect the event-driven nature of the algorithms and the way they process events, easing future translation to neuromorphic hardware. The framework integrates drivers to communicate with the DVS, the DAVIS, the Opal Kelly ATIS, and the CCam ATIS.

Keywords: asynchronous computation; development framework; event-based processing; event-based sensing; silicon retinas.