*Net updates

CVPR:  On the 20th, we turned in a CVPR paper on camera selection for localization.  It was largely based on Mike’s thesis work, but added a distance error plot based upon real data and a “lookup table” approach to modeling the error function.  It got turned in time, but without sufficient editing, though we didn’t really start it until about a week before it was due.

COPRNet: Met with Cris on the 12th at her home and with Kevin Lafferty.  Kevin’s interests are in the spread of parasites, and he did a survey of birds in a specific area over time with a cassette video camera in a large, transparent container and compared it to snail populations.  His advice for us was largely based upon developing inexpensive camera technology that can be “dropped” into remote location, then transmit the data back without need maintain it very regularly.  Our conclusions from it were that we need to ensure we can get good resolution images of the birds and that that is the only way to reliable identify them.  Additionally, we should focus on the identifying white egrets like the great white egret and the snowy egret, which are perennial birds.

We also looked at the path on the opposite side of the slough.  There’s a path that’s accessible either from the beach or the golf course.  Neither are that easy to get to by car.  Much of the path hugs the Slough, with several benches for watching birds.  Though the area near the bridge is not accessible to us, and we’ll have to talk to Cris about getting into there.  The hope is that somewhere along there will give us more interesting locations to put cameras and have a better multiple camera network.

KirbyNet: One node will go in KITP (Kohn Hall) in an office almost right above the building’s main entrance.  Carter and I went over to talk to the person in that office, and found that it would work great.  We have to get it in there next week.

I also talked to the KITP Information Systems Director today about hooking into their network.  He’s concerned about us too much bandwidth, so he would like to know our estimate on how much data we’ll be transferring and when.  Other things on “his” network are backups at night and general usage during the day.

HFHNet: Prof. Manjunath mentioned today about setting up a camera network withing HFH, observing all of the entrances and exits to each floor, so that we can know how many people are on each floor.  I need to make a schematic by Monday, as they’re meeting on Tuesday.  I wish building floorplans were easier to find.