refining infodump

Have a 14 page infodump version of the RE writeup, and have polished the first 4 pages.

SIFT flow

Got SIFT flow working, producing Euclidean and FPLBP distances. Since we need to process lots of files in lots of configurations, I wrote a controller in Scala. Just kicked off a job to get ROC curve data for all easy conditions matched to pose-2 and easy otherwise. SIFT flow takes maybe 20s per image pair, so I'm going to pick it up again tomorrow.

face.com performance characterization and DB class

Spent today getting stuff in order so that face.com's performance on the Comcast MURI dataset can be systematically characterized. Most of the time went into building an HDF5 database storing a bunch of metadata for each image, including the face.com information. The data for the database was stored across a bunch of directories and file formats, so it took me a few attempts to get it constructed. Now that the database exists, I'm writing a script to automatically generate detection and miss examples by condition, including marking up the example image.

Also spent maybe 80min on db-class.org.

mostly moved

For the last couple days, spent most of my time moving and doing stuff related to moving. Dealing with real life sucks; wish I could hire a mom. Also looked at new MURI dataset and did a bit of work on research exam outline.

Google, research exam

Haven't updated this in a while; but hoping to get on it again.

Googled today (third to last day!). Also finished phase 3 of my research exam literature review and emailed Gary and David about dates.

Literature review

Published summaries of three papers, and added many more to the Mendeley group.

Literature review

Summarized a paper and the relevant content in a book.