Rendered at 17:01:18 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
jdhwosnhw 12 hours ago [-]
Fyi the unscented Kalman filter is both easier to implement than the EKF, and also avoids several of the requirements that come along with the need to linearize (such as the differentiability requirement mentioned in the article). Also (to me, at least) the UKF is conceptually much cleaner, as the whole point is to place the approximation in the parameterization of the distribution, rather than on the function operating on that distribution.
Pilling on to say well done on the interactivity and visuals / design overall. I'm working to make producing posts like this universally accessible (http://motate.app/) and posts like yours are an inspiration.
13 hours ago [-]
nickcw 9 hours ago [-]
Great demo! Very interesting to see that if you wiggle the hunter (1) back and forth the accuracy improves.
I think this is expected but interesting to see as you see humans and animals doing exactly this to better gauge how far away something is.
The accuracy also improves (but not as much) if you wiggle the target (2) back and forth which I wasn't expecting.
jmux 17 hours ago [-]
nice work! the interactive visuals are really cool
noen 16 hours ago [-]
Really well written article, thank you!
treycluff 17 hours ago [-]
I love a blog post with interaction
rayhanadev 15 hours ago [-]
love seeing purdue hackers folks on hackernews :)
kritr 14 hours ago [-]
I was going to say the same thing.
Purdue Hackers has grown into a much needed space at Purdue, nice to see the effects going beyond.
https://groups.seas.harvard.edu/courses/cs281/papers/unscent...
I think this is expected but interesting to see as you see humans and animals doing exactly this to better gauge how far away something is.
The accuracy also improves (but not as much) if you wiggle the target (2) back and forth which I wasn't expecting.