Research

The MI³ Lab conducts state-of-the-art research in artificial intelligence, machine learning, computer vision, and biomedical imaging informatics. We focus on developing methods for object recognition, image and video segmentation, image and video synthesis, computer-assisted diagnosis and intervention, 2D image to 3D scene reconstruction, registration/matching, and motion tracking.  More broadly, our research is in the areas of artificial intelligence and data science for healthcare and biomedicine, biomedical informatics, computer graphics, visualization, and human-computer interaction. Research in the lab has been funded by the NIH, NSF, DOE, the Howard Hughes Medical Institute, and the Pennsylvania state.

Research theme 1:

Develop robust image analysis methods and build autonomous and intelligent systems that integrate algorithms with efficient, application-specific designs to solve computational problems in biomedicine, perception, and cognition.

Design intelligent vision algorithms and systems that are capable of learning effectively and reasoning about multiple sources of information in order to achieve functions typical of human vision

Research theme 2:

Create robust medical imaging software based on computer vision and machine learning algorithms that aid medical doctors in accurate and reproducible diagnosis, and help them better understand the anatomical and physiological relationships in normal and diseased states.

Develop image processing and analysis software for analyzing biological images to help biologists and biophysicists in understanding and modeling complex biological pathways and systems.

Diffusion models for image and video generation

coming soon…

GAN models for multimedia content generation

coming soon…

FeelReel: An Interpretable Multimodal Benchmark Dataset with Rich Attributes for Emotion Understanding

coming soon…

Synthetic Data Augmentation

coming soon…