Advanced search

Message boards : News : e-Science 2013 presentation

Author Message
Travis Desell
Volunteer moderator
Project administrator
Project developer
Project scientist
Send message
Joined: 16 Jan 12
Posts: 1813
Combined Credit: 23,514,257
DNA@Home: 293,563
SubsetSum@Home: 349,212
Wildlife@Home: 22,871,482
Wildlife@Home Watched: 212,926s
Wildlife@Home Events: 51
Climate Tweets: 22
Images Observed: 774

              
Message 1939 - Posted: 23 Oct 2013, 6:32:57 UTC

Hi Everyone!

I just finished giving my talk on Wildlife@Home at e-Science 2013. Our project seems to be really well received!

Here's the presentation if you'd like to take a look at the slides:

http://people.cs.und.edu/~tdesell/talks/2013_october_23_escience/index.html

--Travis

Profile Defender
Avatar
Send message
Joined: 4 May 13
Posts: 103
Combined Credit: 1,607,715
DNA@Home: 26,301
SubsetSum@Home: 21,008
Wildlife@Home: 1,560,405
Wildlife@Home Watched: 1,089,206s
Wildlife@Home Events: 343
Climate Tweets: 4
Images Observed: 8

          
Message 1940 - Posted: 23 Oct 2013, 8:31:34 UTC - in response to Message 1939.

Nice presentation. I like the slide about feature detection, that has been something I never understood before.

Travis Desell wrote:


Our project seems to be really well received!


Can you explain it a little more? It would be interesting to know what your audience thought about it in detail.
____________
Proud member of SETI.Germany

Travis Desell
Volunteer moderator
Project administrator
Project developer
Project scientist
Send message
Joined: 16 Jan 12
Posts: 1813
Combined Credit: 23,514,257
DNA@Home: 293,563
SubsetSum@Home: 349,212
Wildlife@Home: 22,871,482
Wildlife@Home Watched: 212,926s
Wildlife@Home Events: 51
Climate Tweets: 22
Images Observed: 774

              
Message 1941 - Posted: 23 Oct 2013, 8:40:03 UTC - in response to Message 1940.
Last modified: 23 Oct 2013, 8:40:22 UTC

Nice presentation. I like the slide about feature detection, that has been something I never understood before.


You can read more about that (and get references to other papers if you're really interested) in the eScience paper I put up awhile back:

http://people.cs.und.edu/~tdesell/papers/2013_escience_wildlife.pdf


Travis Desell wrote:


Our project seems to be really well received!


Can you explain it a little more? It would be interesting to know what your audience thought about it in detail.


Some particular comments were that it would be very interesting to open up our database of videos (along with your observations) to the computer vision community, as a challenging data set for them to work on. We actually are planning on doing this (and it's part of the NSF grant we were recently awarded).

There was some interest in combining audio analysis with the video analysis -- which is something I'll talk to Susan about. In future nesting seasons we could potentially gather audio as well, which might aid in determining what's happening in the video.

Another interesting discussion I had with some conference goers was on coming up with better ways to improve the accuracy of your observations, and gathering information about how the observations are made to come up with metrics on what user activities lead to more accurate observations. A couple ideas were to maybe swap between species and/or locations after some amount of views (to keep things interesting and prevent burn-out or too much automated clicking), and looking into things like how much does using the scroll bar/fast forward/fast reverse correlate to making correct observations -- we could use this information as an aid to determine if anyone is just spamming buttons trying to gain credit.


Post to thread

Message boards : News : e-Science 2013 presentation