|
Greg Hager
John Hopkins University
Functional Languages Meet Vision and Robotics: FRP in Action
Functional Reactive Programming (FRP) was pioneered by Conal Elliot
and Paul Hudak as a way of specifying computations where
time-flow is a central concept. FRP extends a host language with
signals whose values evolve in time. Until recently the only host
language for FRP was Haskell, as originally developed in a computer
animation system called FRAN. Over the last few years we have been
experimenting with these language ideas in two new domains: robotics
and computer vision. The results are two new FRP based languages,
FROB (Functional Robotics) and FVision (Functional Vision).
In this talk, I will discuss the basic FRP framework and how it has
evolved to encompass vision and robotics. Along the way, I'll
describe some of the lessons we've learned about using FRP for
cross-domain integration, and a few of the domain-independent language
abstractions that have emerged. I will also discuss a new C++
implementation of FRP that we have recently developed. We have
implemented a simple vision-guided navigation system in FRP/Haskell
and FRP/C++. I will compare and contrast some of the pros and cons of
both systems.
|
