Multi-reference alignment and clustering

Wednesday, August 16, 2017 - 10:00am - 10:45am
Lind 305
Amit Singer (Princeton University)
Multi-reference alignment consists of estimating a signal from its many noisy shifted copies, where the shifts are unknown. The number of observations needed for accurate estimation, efficient recovery algorithms, extension to recovery of multiple signals, and application to cryo-electron microscopy and X-ray free electron lasers (XFEL) will be discussed.