Blätter-Navigation

Offer 302 out of 586 from 17/06/22, 13:57

logo

Helm­holtz-Zen­trum Dres­den-Ros­sen­dorf e.V.

Through cut­ting-edge research in the fields of ENERGY, HEALTH and MAT­TER, Helm­holtz-Zen­trum Dresden-Rossen­dorf (HZDR) solves some of the press­ing soci­etal and indus­trial chal­lenges of our time. Join our 1.400 employ­ees from more than 50 nations at one of our six research sites and help us mov­ing research to the next level!

Hoff­mann lab at the Insti­tute of Radi­ation Phys­ics is research­ing machine learn­ing based digital twins of laser-driven plasma accel­er­at­ors con­trib­ut­ing to our fun­da­mental research about the­ory, exper­i­mental val­id­a­tion and auto­matic con­trol of this sys­tem enabling prom­ising applic­a­tions like next gen­er­a­tion X-ray light sources. More details about our research can be found here: https://photon-ai-
research.github.io

You will be research­ing machine learn­ing meth­ods that are solv­ing inverse prob­lems encountered by our exper­i­ment­al­ists. From a math­em­at­ical per­spect­ive, you will be deal­ing with the phase prob­lem which is quite prom­in­ently encountered in ana­lysis of X-ray dif­frac­tion data that we are col­lect­ing at our HIBEF instru­ment at EuropeanXFEL (https://www.xfel.eu/index_eng.html). You will be design­ing method that integ­rate inform­a­tion from many ortho­gonal obser­va­tions of the sys­tem’s state, incor­por­ate prior know­ledge about phys­ical pro­cesses while being robust to out-of-dis­tri­bu­tion data. You will be also coordin­at­ing your efforts with all involved parties to stim­u­late fast integ­ra­tion of your code into exper­i­ment at HZDR as well as EuropeanXFEL.

Does that sound inter­est­ing to you as a Phys­i­cist, Math­em­atician or Com­puter Sci­ent­ist? Then sub­mit your applic­a­tions as Postdoc (f/m/d) Inverse Prob­lems/Ima­ging/Optim­isa­tion as soon as pos­sible.

The pos­i­tion will be avail­able from 1 August 2022. The employ­ment con­tract is lim­ited to 31 Decem­ber 2023 but exten­sions will be pos­sible upon suc­cess.

Postdoc (f/m/d) Inverse Prob­lems/Ima­ging/Optim­isa­tion

Working field:

  • Research and imple­ment­a­tion of ML-assisted algorithms for solv­ing ill-posed inverse (ima­ging) prob­lems
  • Pro­act­ive col­lab­or­a­tion with exper­i­ment­al­ists (f/m/d) at HZDR and EuropeanXFEL (Link) to trans­late know­ledge and require­ments for data acquis­i­tion and in-situ data ana­lysis into exper­i­ment
  • Pub­lic­a­tion of papers in high-qual­ity journ­als in Phys­ics, Math­em­at­ics and Machine Learn­ing con­fer­ence
  • (Joint) super­vi­sion of PhD and Mas­ter's stu­dents (f/m/d) work­ing on inverse prob­lems
  • Con­tri­bu­tion to teach­ing basics of inverse prob­lems and ML at uni­versity
  • Act­ively col­lab­or­at­ing with industry, aca­demia, gov­ern­ment labs, and applic­a­tions developers in a vari­ety of ven­ues

Requirements:

  • PhD degree in Com­pu­ta­tional Phys­ics, Math­em­at­ics, Com­puter Sci­ence or a closely related dis­cip­line
  • Good math­em­at­ical back­ground (e.g. lin­ear algebra, stat­ist­ics, optim­isa­tion)
  • Prefer­ably proven know­ledge in solv­ing ill-posed inverse prob­lems (e.g. phase retrieval) and in ana­lysis/recon­struc­tion of X-ray ima­ging mod­al­it­ies (e.g. SAXS, Holo­graphy Pty­cho­graphy)
  • Good pro­gram­ming skills in Python
  • Proven know­ledge of PyT­orch is a strong plus
  • Pub­lic­a­tion record in com­pu­ta­tional ima­ging, inverse prob­lems, and/or machine learn­ing
  • Inter­ested in Plasma Phys­ics, inverse prob­lems, phase retrieval, Machine Learn­ing, Image Ana­lysis and Deep Gen­er­at­ive Mod­els

What we offer:

  • A vibrant research com­munity in an open, diverse and inter­na­tional work envir­on­ment
  • Sci­entific excel­lence and extens­ive pro­fes­sional net­work­ing oppor­tun­it­ies
  • The employ­ment con­tract is lim­ited to to 31 Decem­ber 2023 but exten­sions will be pos­sible upon suc­cess
  • Salary and social bene­fits in accord­ance with the col­lect­ive agree­ment for the pub­lic sec­tor (TVöD-Bund) includ­ing 30 days of paid hol­i­day leave, com­pany pen­sion scheme (VBL)
  • We sup­port a good work-life bal­ance with the pos­sib­il­ity of part-time employ­ment and flex­ible work­ing hours
  • Numer­ous com­pany health man­age­ment offer­ings
  • An employer sub­sidy for the VVO job ticket

How to apply:

Kindly sub­mit your com­pleted applic­a­tion (includ­ing cover let­ter, CV, dip­lo­mas/tran­scripts, etc.) only via our Online-applic­a­tion-sys­tem: https://www.hzdr.de/db/Cms?pNid=490&pLang=en&pOid=66231