| The Advanced Photon Source (APS) at Argonne organizes an ongoing seminar series, the APS Scientific Computation Seminar Series, focused on various aspects of computation for synchrotron science; talks are related to data analysis, reconstruction, theory, simulation, optimization, machine learning, automation, and visualization. The seminar provides an opportunity to present and learn about state-of-the-art software and the application of computational, AI/ML and mathematical techniques to synchrotron science. |
| Title: | Algorithm-Driven Advances for Autonomous Scientific Imaging Instruments | | Presenter(s): | S. V. Venkatakrishnan, R&D Staff Member
Multimodal Sensor Analytics Group, Oak Ridge National Laboratory | | Date: | August 12, 2024 | | Time: | 1:00 PM (Central Daylight Savings Time) | | Location: | Join ZoomGov Meeting
https://argonne.zoomgov.com/j/1601444470?pwd=N1phbHZVdCtmcVR5cGh0c1Zhc0…
Meeting ID: 160 144 4470
Passcode: 937918
One tap mobile
+16692545252,,1601444470# US (San Jose) +16468287666,,1601444470# US (New York) Dial by your location
+1 669 254 5252 US (San Jose)
+1 646 828 7666 US (New York)
+1 646 964 1167 US (US Spanish Line) +1 669 216 1590 US (San Jose)
+1 415 449 4000 US (US Spanish Line) +1 551 285 1373 US
Meeting ID: 160 144 4470
Find your local number: https://argonne.zoomgov.com/u/af2crdvQy
| | Abstract: | Scientific imaging and diffraction instruments are widely used for non-destructive characterization to reveal important material properties such as local lattice parameters, chemical composition, and residual strains. For certain applications such as neutron computed tomography (CT), neutron-diffraction based strain imaging, and high-resolution X-ray CT, the measurement times can be long - spanning the order of hours or even days. Conventional instrument operation involves mounting the sample on a stage, acquiring multiple measurements by orienting the sample in a sequential manner according to a predetermined plan, and processing the data at the end of the experiment using conventional reconstruction algorithms at the end of the scan in order to obtain the information of interest. However, this open-loop approach is not robust to imperfections in the acquisition system and requires a large number of measurements to obtain high-quality 3D information when using conventional algorithms thereby limiting instrument throughput. Furthermore, the instrument users have limited feedback of the overall structures in the sample during the experiment which can potentially lead sub-optimal use of instrument time if the experiment did not reveal the relevant information. In this talk, Venkat will present research on designing algorithms for building autonomous scientific imaging and diffraction instruments - that make the fewest possible yet most informative measurements and adapt to the specifics of the sample to be measured. The approach involves using controllable sample stages, designing new acquisition protocols, moving measurements to powerful compute nodes, developing advanced image reconstruction algorithms and processing the results from partial measurements using machine-learning methods to drive the next set of measurements. Using examples from hyper-spectral neutron CT, X-ray CT, and neutron diffraction Venkat will demonstrate how the methods we have developed can reduce overall measurement times dramatically while providing useful real-time feedback to end users. | | Bio: | S. V. Venkatakrishnan received the Ph.D. degree from the School of Electrical and Computer Engineering at Purdue University in 2014. He is currently a R&D Staff member at the Multimodal Sensor Analytics group at Oak Ridge National Laboratory developing computational imaging algorithms in support of the lab’s efforts in ultrasound, X-ray, electron and neutron-based systems. His research interests include Computational Imaging, Inverse Problems and Machine Learning. He was awarded a Presidential Scholar Award at the Microscopy and Microanalysis conference (2014), and the Eugene P. Wigner Distinguished Staff Fellowship at Oak Ridge National Lab (2016-2019). Dr. Venkatakrishnan is a senior member of the IEEE and has served as an Associate Editor of the IEEE Trans. On Computational Imaging (2018-2024). |
|
