Efficient detection of objects in images is complicated by variations of object appearance due to intra-class object differences, articulation, lighting, occlusions, and aspect va...
In this work, we develop and evaluate a wide range of feature spaces for deriving Levinstyle verb classifications (Levin, 1993). We perform the classification experiments using Ba...
We present a unified model for face detection, pose estimation, and landmark estimation in real-world, cluttered images. Our model is based on a mixtures of trees with a shared p...
In this paper, we propose a novel technique for modelbased recognition of complex object motion trajectories using Gaussian Mixture Models (GMM). We build our models on Principal ...
Faisal I. Bashir, Ashfaq A. Khokhar, Dan Schonfeld
The use of the statistical technique of mixture model analysis as a tool for early prediction of fault-prone program modules is investigated. The Expectation-Maximum likelihood (E...