4/21/2021 Laura Schmitt, Illinois CS
CS faculty Girish Chowdhary, John Hart, Derek Hoiem, and Edgar Solomonik are receiving college- and campus-level awards this spring for their research and instructional accomplishments.
Written by Laura Schmitt, Illinois CS
CS faculty Girish Chowdhary, John Hart, Derek Hoiem, and Edgar Solomonik are receiving college- and campus-level awards this spring for their research and instructional accomplishments.
A leading researcher in robotics and autonomous systems, associate professor Girish Chowdhary has won the 2021 Paul Funk Faculty Award for Excellence in Research from the Illinois College of Agricultural, Consumer and Environmental Sciences. He won a comparable faculty research award from the Illinois Grainger College of Engineering in 2020.
“I am greatly honored and humbled to receive this award,” said Chowdhary. “A key application of my group’s research in field-robotics is to create new options for farmers in the form of small mobile robots. We are pursuing novel techniques that bring together robotics, AI, and agricultural engineering. This recognition from ACES of the progress we are making towards this goal is highly encouraging, and exemplary of the interdisciplinary work that Illinois is known for.”
Chowdhary, who holds a joint appointment with computer science and agricultural and biological engineering (ABE), has an impressive track record of creating collaborating teams of autonomous field robots for agricultural, defense, exploration, and other outdoor operations. These robots incorporate AI, learning, estimation, and control algorithms in order to solve real-world challenges in harsh, uncertain, and constantly changing environments.
In recent years, he is perhaps best known for his creation of the TerraSentia robot, a compact and teachable class of robot that uses a number of sensors to collect data on crop health, as well as machine learning-based analytics to convert this data into actionable insights for farmers. TerraSentia robots autonomously move through rows of corn under the plant canopy, where GPS is less effective in guiding their movements.
In 2016, Chowdhary co-founded EarthSense to commercialize the technology underlying the TerraSentia robot. To date, the company has produced and deployed 80 of these robots demonstrating that the small agrobots can be highly rugged and can be produced at highly competitive costs.
It’s feasible that teams of his agbots could reduce the need for herbicides in farm fields. He and his research collaborators have successfully demonstrated through simulation that a team of mechanical weeding robots can effectively keep a field weed free by using AI to coordinate and work together. He is also working with a number of colleagues in CS and ACES on technologies for making it easier and more profitable for farmers to plan cover-crops.
Chowdhary is director of the Distributed Autonomous Systems Laboratory and the Field Robotics Engineering and Sciences Hub (FRESH) in the ABE department. He is also the chief scientist at the Illinois Autonomous Farm that has been established jointly by ABE and the Illinois Center for Digital Agriculture.
In addition to his campus recognition, Chowdhary has received several other honors, including the prestigious Air Force Office of Scientific Research (AFOSR) Young Investigator Research Program award, the Dave Ward Memorial award by Aerospace Control Guidance and Systems Group (ACGSC) for outstanding contributions to Aerospace GNC, and several best paper awards, including a best systems paper award at the Robotic Science and Systems (RSS) conference for his work on the agricultural robot TerraSentia.
Professor John C. Hart has won the 2021 Campus Award for Excellence in Online and Distance Teaching. As both a faculty member teaching online courses and as the department’s director of online education, Hart has made an indelible mark on the development of a professional workforce in computer and data science.
A talented and dedicated educator, Hart conceptualized, designed, and implemented the widely successful University of Illinois online Master of CS in Data Science (MCS-DS) degree track, which has since expanded to enroll students in a more general Master of Computer Science (MCS) degree program. Now in its fifth year, the online MCS currently enrolls over 1,200 students, many of whom are full-time working professionals with bachelor’s degrees in non-CS disciplines.
The Illinois online degree program is helping meet the increasing demand for data scientists across a range of fields, including healthcare, telecommunications, media and entertainment, banking, and insurance sectors.
“This award honors the entire online MCS team who work diligently to provide this educational opportunity to our highly qualified students who have work, family or other obligations that don’t allow them the privilege to move to campus to study full time,” said Hart.
One company that is taking advantage of the MCS-DS degree is C3.ai, a leading enterprise artificial intelligence software provider for accelerating digital transformation. Founded by CS alumnus Thomas M. Siebel (BA History ’75, MBA ’83, MS CS ’85), C3.ai pays the tuition for its employees who earn the degree, and then retains them with a salary increase, a cash bonus and stock options.
Hart also developed the modern equivalent of an influential textbook when he launched a Data Visualization massive open online course (MOOC) on the Coursera platform in 2016. To date, this course has reached more than 350,000 worldwide learners—for anyone wanting to gain insight into their data—and has achieved a 96 percent approval rating.
Hart’s influence extends beyond the University of Illinois community. Michael Gleicher, a CS professor at the University of Wisconsin-Madison, uses materials that Hart has developed in his own computer graphics course instruction.
“While many people prepare materials for classes, John’s is unique in that he finds ways to convey his enthusiasm for the field, creativity, and personality into the documents,” Gleicher said. “Through the use of hand-drawn comic book-style illustrations, examples, and clear (but not overly colloquial) prose, he is able to convey complex mathematical concepts in an inspiring way. He makes his passion and love of the topic emerge even when introducing equations.”
Associate professor Derek Hoiem has won a 2021 Grainger College of Engineering Dean’s Excellence in Research Award. An expert in computer vision, Hoiem has made important and lasting research contributions to the areas of 3D scene understanding, object recognition, and task representation and learning.
In the area of 3D scene understanding, Hoiem’s early work was the first to demonstrate an ability to create 3D scenes from single images by “recognizing” geometry. Recently, he has published many well-cited papers on recovering indoor scene layout from panoramas and object shape from images, as well as new fiducial marker designs and improved methods to recover 3D models from multiple images.
Hoiem’s contributions to the area of object recognition include using scene geometry to improve detection accuracy, attribute-based representations, and category-independent object region proposals. His work on attributes helped spawn sub-areas of attribute recognition and zero shot learning.
The specific techniques he developed for object proposals were replaced by simpler and faster methods, but the underlying idea was important for development of early deep network-based object detectors such as RCNN. His tools to diagnose error in object detectors continue to be widely used to characterize performance.
His recent work on object recognition has focused on the intersection of vision and language, with methods to answer questions based on images, to learn word representations from image annotations, and to generate cartoons based on scene descriptions.
In 2016, Hoiem co-founded Reconstruct, a remote construction progress monitoring firm, to commercialize technology that integrates 3D models from images, building plans, and schedules. Reconstruct shows “what is there” vs. “what was planned.”
As CTO, he led the software development team that built Reconstruct’s web platform that has been used by more than 40 companies on 400+ construction and inspection projects worldwide. In 2019, he transitioned to chief science officer, and he continues to lead the computer vision group in advancing state-of-the-art in 3D reconstruction and automated progress monitoring.
Hoiem is co-PI and part of an interdisciplinary team of Illinois researchers who recently received funding from both the National Science Foundation and the Discovery Partners Institute to advance the use of AI in construction.
A previous recipient of a Dean’s Excellence in Research award for an assistant professor, Hoiem has received other Illinois engineering and external awards, including a 2015 Campus Distinguished Promotion Award, 2014 CS Department C.W. Gear Junior Faculty Award, 2014 IEEE PAMI Young Researcher Award, and 2011 NSF CAREER award.
“It’s a great honor to be recognized among so many amazing faculty at Illinois,” said Hoiem. “I’m super excited about our current research directions in general purpose vision and mixed view geometry, which will help computer vision and AI make an even greater impact. It’s wonderful to have the opportunity to collaborate with such talented students and researchers.”
Assistant professor Edgar Solomonik received a 2021 Grainger College of Engineering Dean’s Excellence in Research Award. Since joining the Illinois faculty in 2016, Solomonik has been conducting innovative research in the design of efficient numerical algorithms for tensor computations in order to minimize communication costs in high-performance computing.
Tensor computations are a class of challenging calculations characterized by a large number of parameters or unknown variables. Tensors provide a mathematical framework for solving complex and data-intensive problems in physics, chemistry, and bioinformatics.
In addition to this college award, Solomonik has earned several other external research awards, including a prestigious NSF CAREER Award ($500,000) to further develop a high-level tensor library (Cyclops Tensor Framework) that will accelerate analysis of large graphs, approximation of multidimensional datasets by tensor decompositions, and simulation of quantum systems—the latter being a major focus of his research group. Ultimately, researchers can deploy the tensor-based techniques from this library on massively parallel computing systems, making new innovations in computational science possible.
In fact, scientists at IBM and Lawrence Livermore National Lab used Cyclops to perform the largest-ever quantum circuit simulation in 2019.
The impact of Cyclops extends to quantum chemistry, where complex methods that require tensor algebra are trivially expressed in high-level Cyclops syntax. Solomonik is among four principal investigators from Illinois, Caltech, and Columbia who received $1.9 million in NSF Cyberinfrastructure for Sustained Scientific Innovation (CSSI) funding to support this work.
Among his other research-related awards are the 2020 Society for Industrial and Applied Mathematics (SIAM) Activity Group on Supercomputing’s Early Career Prize, the 2018 IEEE Computer Society’s TCHPC Award for Excellence for Early Career Researchers in High Performance Computing, and the 2016 Alston S. Householder Prize for the best dissertation in numerical linear algebra (awarded every 3 years).
Solomonik is a member of the Grainger College of Engineering’s Illinois Quantum Information Science and Technology Center (IQUIST).