Had a meeting with the group. Went over the classification that I have been doing, separating out about 10 classes of items to then do more analysis on the structure features. Linda didn't feel that we should continue on that path, but should just use color, texture, and structure analyses, combined with kmeans clustering, to then try to classify images. Sal talked a lot about wanting consistent segmentation among the color region analysis from Yi's program, but Linda informed him that he could never get consistent segmentation of images. Went back and forth for a while. I explained the idea to everyone on the board to make sure that I got it, since most of it has to do with my implementation. Viet will be finishing up a generic version of kmeans classification for me to use.

Met with Oren today to discuss job prospects. He said that he didn't have anything for me, but would forward my resume to some people in LA. Emailed Linda about this, hoping to extend our contacts to the other people who she knows. Must continue the job search more tomorrow, as she seems to have left for the day.

Talked to Linda about sending my resume to more people. Did some of my own work to get that ball rolling.

Commented most of my code so that Sal, or anyone else, should be able to basically understand what is going on. Extending the processing for the color and structure attributes.

Today I wrote some more code, but dealt with a loss in the family for one of my officemates. She came in later in the afternoon and I tried to relax her. I think we may have left early on this day.

Basically did something similar to yesterday.