Seeing reality anew. Revolutionizing object detection and anomaly detection in computer vision.

Dr. Michael J. Jones was elevated to IEEE Fellow, as recommended by the IEEE Computer Society, for contributions to the computer vision topics of object detection and video-based anomaly detection. Detecting different types of objects (such as faces or people) in an image is a fundamental problem in computer vision and is the starting point for many image processing applications. His research demonstrated a new way of detecting objects that is both accurate and very fast. Mike has also made important contributions to the problem of detecting unusual activity from video of a scene, which has important applications in surveillance and security.

Mike's early work with colleague Dr. Paul Viola led to the development of the first real-time face detector that ran on a standard CPU. The Viola-Jones face detector demonstrated the power of the Adaboost algorithm for machine learning applications to the computer vision community.

Adaboost is used to automatically find the most accurate set of image features for detecting a particular class of objects, such as faces. This work also introduced the idea of the integral image representation which makes the computation of certain simple image features extremely efficient. Finally, his work showed the power of a cascade of increasingly more complex classifiers which greatly speeds up object detection. Together these ideas created a new paradigm for fast and accurate object detection in images.

Mike's more recent work on video-based anomaly detection introduces new methods for efficiently modeling a dynamic scene from video that shows normal activity. Unusual activity in the scene can then be detected by noticing differences from the normal model. This technology frees humans from the tedious job of monitoring surveillance video for safety and security purposes.

Mike's elevation to IEEE Fellow "helps to shine a spotlight on the world-class research being done at MERL," he states. "It shows the key role that an industrial research laboratory like MERL can play in advancing technology for the world.“

Speaking about what's next, he says, "In the future I hope to continue finding new solutions to difficult research problems and to mentor students to help them find success doing research in computer vision and machine learning."