Interesting tests done by the UNC Gamma Group about doing lookahead searchs with collision avoidance algos youtube.com/watch?v=I09heZ… #navigation
5:26pm Jan 14th 2013 via TweetDeck
@clodericmars Interesting results, but do you have any data on how "realistic" it is? It seems a bit too optimized to me.
8:48am Jan 15th 2013 via Netvibes Widget (deprecated) in reply to clodericmars
@LarrxX I didn't read the article in details yet (gamma.cs.unc.edu/lookahead/) but it seems they didn't consider measuring the fidelity yet.
8:51am Jan 15th 2013 via TweetDeck in reply to LarrxX
@clodericmars But what do YOU think? You've got more experience than I in this particular field :)
8:54am Jan 15th 2013 via Netvibes Widget (deprecated) in reply to clodericmars
@LarrxX The test scenarios are all but realistic in the first place, at the agent level it looks mostly better than without lookahead.
9:00am Jan 15th 2013 via TweetDeck in reply to LarrxX
@LarrxX i'd like to see what it could achieved in very dense scenariis such as the one we worked on at @golaem in trains.
9:01am Jan 15th 2013 via TweetDeck in reply to LarrxX
@clodericmars @LarrxX we'll run a test with the current Extrapol algorithm to compare with the video results
9:09am Jan 15th 2013 via TweetDeck in reply to clodericmars
@michaelrouille Keep us informed! Golaem FTW! :-D
9:51am Jan 15th 2013 via Netvibes Widget (deprecated) in reply to michaelrouille
Merge in another conversation (paste in a tweet URL):
Or add a single tweet to the end of the page:
Connect with Twitter
A website by @jwheare