bonvision / benchmarks

Various performance and latency benchmarks for the BonVision package

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

BonVision Benchmarks

This repository contains various performance and latency benchmarks for the BonVision package. Stimulus generation, data acquisition, and analysis scripts used to benchmark BonVision against PsychoPy and PsychToolbox are provided.

How to use

All data acquisition is done using the Harp behavior board, which can be obtained at: https://www.cf-hw.org/harp/behavior

Drivers and runtime dependencies for the Harp stack can be found at: https://bitbucket.org/fchampalimaud/downloads/downloads/

Scripts for closed loop latency and frame-rate measurements were developed using the Bonsai visual programming language, and are executed independently of stimulus generation. A self-contained bootstrapper executable can be found in the bonsai folder in this repository. Running the executable should automatically reproduce the environment configuration used for data acquisition.

Closed loop latency benchmarks were additionally automated using an Arduino Uno generating a pseudo-random sequence of activations to the Harp behavior board. Both input activations and photodiode display responses were recorded with a common timestamp generated by the Harp board. For measuring head-mounted display latency, an infrared cutoff filter was introduced between the photodiode and the display to remove strong periodic infrared light pulses generated from the headset itself.

About

Various performance and latency benchmarks for the BonVision package

License:MIT License


Languages

Language:MATLAB 46.7%Language:Python 40.4%Language:C# 9.2%Language:C++ 2.0%Language:PowerShell 1.5%Language:Batchfile 0.3%