Intelligent Camera Networks

Networks of cameras are nowadays installed in various places. However, most of the data collected by such networks are analyzed manually, thus severely limiting their application. In this research, which is divided into different sub-projects, we are studying the core scientific and technical issues related to camera networks, such as information fusion, network control, activity analysis, the interplay between networking and computer vision, and computational complexity. More...

Wide Area Scene Analysis

Video-based analysis of large areas and long time periods is a very challenging problem because of the need to maintain long-term stability of the tracks and understand the spatio-temporal behavior patterns. In this research, we will be developing algorithms for tracking and activity analysis in camera networks covering a large area. More...

Activity Recognition and Event Search in Video Databases

The goal of this project is to retrieve segments from a video database given a video clip of an activity of interest. Not only does this require development of activity recognition algorithms, these algorithms need to be integrated with video search and retrieval methods in large databases. More...

From Physics-based Models to Statistical Learning

The goal of this project is to develop a theoretical framework for understanding the various factors that affect image formation, e.g., object shape, motion, surface reflectance properties, lighting conditions and camera parameters, and use it for modeling and recognition of in video, in combination with statistical learning approaches. More...

Video-Based Face Recognition

Face recognition continues to be one of the important problems in computer vision. Our goal is to develop video-based face recognition algorithms that are robust to changes of pose and lighting. More...

Image Analysis for Modeling Cell Growth Dynamics

The goal of this project is identification, spatio-temporal modeling and recognition of dynamical patterns inherent in developmental biology through the use of novel computational tools in image analysis, statistical data aggregation, pattern recognition, machine learning and dynamical modeling. This will lead to the development of new methods for addressing some outstanding challenges in this area, i.e., computing cell lineages, identifying long-term patterns in the tracked output and learning functional models of the dynamics of cell growth and division. More...

Query-Based Search of Biological Image Databases

A fundamental and hard question in biology is identification: given a biological sample, we want to identify its closest match among a set of known samples. In this project, we focus on nematodes, which are a very important group of animals. A key problem is that nematodes are particularly difficult to identify: the average identification currently takes approximately 2 days, and a researcher needs 3-5 years of effort to obtain “fluency”. One of the goals of this project is to develop image analysis methods for identification of nematodes. More...

VideoWeb: A Video Network Lab

This unique facility will consist of about 80 video cameras, a few infra-red cameras, and a collection of acoustic, seismic and vibration sensors connected through a wireless network in an outdoor environment. This will enable researchers to work on the problems at the intersection of a number of diverse disciplines, like computer vision and image processing, networking, communications, databases, and others. More...

Imaging and Non-imaging Sensor Fusion

This laboratory will consist of a range of sensory devices to enable research on urban disaster management by collecting, integrating and analyzing data from video and other sensors. The integration of these sensors will lead to a more holistic approach to understanding the disaster environment, and provide better decision support systems for emergency personnel. More...

Multi-terminal Video Compression