Solving today’s biggest engineering challenges requires innovation, creativity and an appreciation of when new methods may work better than what was done in the past. This is why IUPUI’s Lauren Christopher has made artificial intelligence such an integral component of her current research.
Christopher’s interest in AI stems from her experience in imaging. As an engineer at RCA, she did a lot of work with image compression and realized that 3D and the ability to view objects in a visual scene was the next step.
“If a human views a particular scene, they may see a gray wall and a desk with a computer on it, but I don’t have to tell you that this pixel is grey and that pixel is grey, which is what we were doing in the past,” Christopher said. “When looking at object detections in a scene, it is clear that AI was looking at it more like a human would than a camera would. That is what I like about AI -- not only does it work and is interesting, but it functions similar to how a human brain does.”
Christopher is currently engaged in multiple projects that draw on her knowledge and experience in using AI for object detection, and one area of focus is automotive safety. The Indiana Department of Transportation provided a grant to analyze traffic patterns in Indianapolis, and she is working with PhD student Mei Qiu within a larger project of IUPUI’s Transportation and Autonomous Systems Institute that uses AI to detect cars in a video scene, data that provides a live count of the cars.
Christopher is also applying her knowledge of object detection to national defense. During a sabbatical in 2020, she worked with the U.S. Navy at the Naval Surface Warfare Center, Crane Division on developing an automated system to score how well certain countermeasures – such as flares that are used to deflect surface-to-air missiles from hitting its target – performed against attacks.
“In a war zone, people may use missiles to try to take down helicopters and planes, and countermeasures, like flares, that deflect the bad guys’ missiles save lives,” said Christopher, who spent time observing the testing at the White Sands Missile Range. “It is currently a very human-intensive process to review testing video and mark whether a missile was deflected by these countermeasures or not. The AI allows an automated scoring system, which helps by only requiring humans to review anomalies that required additional attention.”
Last year, Christopher also embarked on a two-year Naval Imagery Infrastructure Revitalization grant to develop models that will help answer a question that often came up in discussions during her sabbatical: how close is virtual world data to the real world?
“To gain a better understanding of computer generated versus real world, we’re digging into the neural net guts to see what’s going on and how it is representing those images,” Christopher said. “We’re understanding how those mathematical models may be similar to how humans do things.”
Christopher is also engaged in a project with Dr. Dawn Neumann, an associate professor of physical medicine and rehabilitation at the IU School of Medicine, who works with patients who have traumatic brain injuries using the Theory of Mind – a cognitive skill involving the ability to think about your own or others’ mental states, like emotions and desire, and how to attribute them.
Neumann’s patients watch a sequence of videos and then answer a series of questions. During this process, Christopher uses AI to track where a patient’s eyes are looking on the screen. If a person is not looking at the correct places on a person’s face, they may have trouble interpreting others’ expressions and whether they are telling the truth.
“We’re trying to predict from the eye tracking how they are going to answer those questions, so we break it down using AI to see where the facial landmarks are and then use AI to decide or predict how the person will answer based on where their eyes are,” Christopher said. “Knowing this information can help with the rehab process, and it can even be used for stroke patients or others where there is some disconnect between what you think you’re seeing and what you are actually seeing.”
Christopher said one of the neatest applications she has been involved with is an imaging project with a recent PhD graduate, David Emerson. This project uses AI with the concept of depth from defocus, an algorithm using in-focus and out-of-focus image pairs to determine depth without using radar. She said while the depth from defocus idea has been around for a while, she is not aware of any other researchers who have applied the use of AI that develops the solution for the depth.
There is no doubt that Christopher’s career as a woman in engineering has been a success. Even before coming to IUPUI, she led the first DIRECTV receiver design and the first DVD player design as part of the Thomson Consumer Electronics team. In fact, she was inducted into the Consumer Electronics Hall of Fame in 2010 for her work leading the development of the DIRECTV set-top box.
“Being a woman in science and engineering has always been good for me, and I was helped by the prevailing wisdom in the 60s and 70s that women can do anything,” Christopher said. “Cultural norms may shift but being a tech person is not considered geeky anymore, thankfully.”
Christopher believes it is important to keep the next generation of women in science engaged and excited about opportunities that exist.
“We have to keep reminding middle school and high school girls that technology is cool and not just for boys,” Christopher said. “IUPUI does a great job at encouraging girls to get involved in these fields.”
Research Impact resources and social media channels