Monitoring Space Without Sacrificing Privacy – CURF Introduction: Friedrich Doku

My name is Friedrich Doku, and I am an undergraduate at the University of Pittsburgh, majoring in computer science. This spring, I am working with Dr. Adam J. Lee to develop a system that collects information about a space without invading the privacy of individuals. Our approach to this problem is designing a video processing system where video frames are processed inside a Trusted-execution environment (TEE). The system collects useful data and ignores any data that would cause an invasion of an individual’s privacy. I call this system XPhoton. The data inside of a TEE cannot be exported or accessed from the outside (unless permission is granted explicitly). XPhoton places its images inside its TEE, so any other program cannot access the data, thus, providing the video frames with confidentiality.

 XPhoton’s goal is to allow monitoring of space while keeping the raw video contents private. For example, say you want to count the number of people in a room without watching the video. XPhoton will run an image recognition algorithm that will tell you how many people are in the room while keeping the video content protected in the TEE.

Today, we are surrounded by many video recording devices, from security cameras to smart home cameras, such as Nest cameras. Many of us are so used to this that we never think about whether we can trust these devices. Security cameras capture more than what they need to provide security. Security cameras are for keeping people and property safe, but they also create temptations to abuse them for personal purposes. Furthermore, having humans watch the uninterpreted video feed can lead to many biases, such as racial profiling. Nest cameras are used to monitor home activity, but these cameras could also be monitoring you as there is no way to control what Nest’s servers do with your video. Unlike XPhoton, these devices don’t protect the raw video content. 

XPhoton’s design makes it perfect for Covid-19 space monitoring. The goal of Covid-19 space monitoring is to monitor how space is being used to help people make decisions that reduce their risk of infection. For example, XPhoton can be set up in a building to provide information about how many people went up the elevator and how many people went up the stairs. If a lot of people went up the elevator, XPhoton would recommend that you take the stairs instead of the elevator because the probability of Covid-19 being in the elevator is much higher. XPhoton would use an image recognition algorithm to determine whether people used the stairs or the elevator more while protecting the video content.

In the future, I want to become a researcher in the field of computer science. I think that the CURF will help me pursue this goal because it provides me with an opportunity to gain real research experience. Research is something that I am excited about because it gives me the freedom to explore new ideas in my favorite areas of computer science.

Leave a Reply