Amir K. Ziabari

Dr. Amir Koushyar Ziabari is a senior R&D staff scientist at the Oak Ridge National Laboratory (ORNL), within the Electrification and Energy Infrastructure Division’s Multimodal Sensor Analytics group. His academic journey includes a tenure as a Postdoc at Purdue University’s Integrated Imaging group, Department of Electrical and Computer Engineering (ECE), where he also completed his PhD in August 2016. Dr. Ziabari’s research has been at the intersection of physics, signal/image processing, computational imaging and machine/deep learning to unravel complex multiscale experimental physics phenomena. His work is pivotal in data analytics, driven by a synergy of data-driven and physics-based methodologies for image reconstruction and segmentation for various scientific imaging applications in domains including advanced manufacturing, medical imaging, nuclear, and materials science.

Dr. Ziabari’s research philosophy succinctly can be characterized as “data science for science”, where he has been leading efforts on developing innovative physics-based computational imaging, computer vision, and machine/deep learning algorithms for scientific imaging applications. These tools are designed to meticulously process and analyze multiscale scientific and medical imaging data, enhancing the capabilities of imaging systems and fostering advancements in image reconstruction, processing, detection, segmentation, and classification. His recent initiatives involve leveraging Generative AI models to synthesize realistic experimental data from physics-based simulations, thereby generating essential multi-modal training data for scientific imaging applications.

Currently leading projects on Advanced Materials & Manufacturing Technologies (AMMT) and DOE  Advanced Materials & Manufacturing Technologies Office (AMMTO) programs at ORNL,  Dr. Ziabari works closely with a dynamic team engaged in developing cutting-edge computational imaging and machine learning algorithms for a variety of scientific imaging applications. Beyond his R&D and leadership responsibilities at ORNL, Dr. Ziabari involves in three major societies—IEEE, ASTM, and OSA—where his recent promotion to IEEE senior member and technical committee member of the IEEE Computational Imaging allows him to more meaningfully contribute to the society.

His recent work in rapid X-ray CT characterization for additive manufacturing, leveraging deep learning-based reconstruction to significantly enhance the throughput and quality of imaging for dense metal parts, has resulted in collaborative efforts and in turn a licensing agreement and partnership with ZEISS, which in turn demonstrates  his commitment to bridging the gap between academic research and practical, real-world applications.

Dr. Ziabari has directly contributed to recruiting, mentoring and the growth of postdocs at ORNL, focused on creating a diverse and inclusive research environment. His vision for expanding research into non-destructive material characterization and multi-modal data analysis aims to further advance materials science applications at ORNL and beyond.

Dr. Ziabari has successfully secured funding as both principal investigator (PI) and co-PI, totaling over $3.5 million. His current and future endeavors will be focused on physics-based Computational imaging and Machine/Deep Learning for multi-modal scientific imaging applications in various scientific domains.