I am interested in developing machine learning, particularly probabilistic models that are simple and intuitive, and facilitate easy understanding of the various aspects of the problem. My main application domain has been computer vision because of the potential for so many useful applications. Apart from computer vision problems, I have applied machine learning to various other domains like text analysis, financial data, and bioinformatics.

In computer vision, I have worked on a wide range of problems, both "low-level" ones such as motion segmentation and "high-level" ones like object detection. Low-level algorithms are interesting to me because these algorithms are currently accurate enough for building reasonably reliable vision systems, for instance automatic surveillance systems. More importantly, they are interesting to me because successful use of low-level techniques can help in developing more accurate systems for higher level tasks like object detection and recognition.

I am interested in object detection and recognition because of the potential for so many applications for users. I have worked on face detection, face pose estimation, face alignment, and object detection.

Prior to joining UMass, I got my Masters at the University of Kansas, where I worked on developing the system and algorithms for automatic tracking of objects in surveillance video. KUVidAnalysis was a great way to get started in computer vision.

Learning and vision apart, one of the most challenging and unique projects I worked on was the Nanosatellite project at Kansas where we developed a satellite system that we launched regularly on high-altitude balloons (pictured above).

The projects page has a more details about my research.


Coursework

PhD

  • Machine Learning
  • Graphical Models
  • Advanced Algorithms
  • Artificial Intelligence
  • Distributed Operating Systems
  • Computer Architecture

MS

  • Computer Vision
  • Digital Video for Multimedia Applications
  • Information Retrieval
  • Natural Language Processing
  • Neural Networks and Fuzzy Systems
  • Mobile Robotics
  • Image Processing
  • Digital Signal Processing