You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
My name is Robert Twyman, I am a Ph.D student in Medical Imaging at University College London, UK. The focus of my Ph.D is accelerating iterative Positron Emision Tomography (PET) reconstruction using stochastic algorithms, such as SAG, SAGA, and SVRG. I have shown that image reconstruction can be accelerated using these algorithms, compared to standard methods used in clinical practice. Furthermore, it has been shown that L-BFGS-B can be used to accelerated convergence at later iterations in the reconstruction process.
I want to bring these two projects together and apply stochastic second order methods to PET image reconstruction. I have been looking at sQN and sLBFGS. I found this github repository while browsing to find avalible python code by the authors but it is only avalible in Julia.
Dear XinEDprob,
My name is Robert Twyman, I am a Ph.D student in Medical Imaging at University College London, UK. The focus of my Ph.D is accelerating iterative Positron Emision Tomography (PET) reconstruction using stochastic algorithms, such as SAG, SAGA, and SVRG. I have shown that image reconstruction can be accelerated using these algorithms, compared to standard methods used in clinical practice. Furthermore, it has been shown that L-BFGS-B can be used to accelerated convergence at later iterations in the reconstruction process.
I want to bring these two projects together and apply stochastic second order methods to PET image reconstruction. I have been looking at sQN and sLBFGS. I found this github repository while browsing to find avalible python code by the authors but it is only avalible in Julia.
I was hoping that you might liscene this repository for use by others, e.g. https://github.com/pcmoritz/slbfgs/blob/master/LICENSE
Regards,
Robert Twyman
The text was updated successfully, but these errors were encountered: