Twin gaussian processes for structured prediction
L Bo, C Sminchisescu - International Journal of Computer Vision, 2010 - Springer
We describe twin Gaussian processes (TGP), a generic structured prediction method that
uses Gaussian process (GP) priors on both covariates and responses, both multivariate, and …
uses Gaussian process (GP) priors on both covariates and responses, both multivariate, and …
Learning joint top-down and bottom-up processes for 3D visual inference
C Sminchisescu, A Kanaujia… - 2006 IEEE Computer …, 2006 - ieeexplore.ieee.org
We present an algorithm for jointly learning a consistent bidirectional generative-recognition
model that combines top-down and bottom-up processing for monocular 3d human motion …
model that combines top-down and bottom-up processing for monocular 3d human motion …
The joint manifold model for semi-supervised multi-valued regression
R Navaratnam, AW Fitzgibbon… - 2007 IEEE 11th …, 2007 - ieeexplore.ieee.org
Many computer vision tasks may be expressed as the problem of learning a mapping
between image space and a parameter space. For example, in human body pose …
between image space and a parameter space. For example, in human body pose …
BM³E: Discriminative Density Propagation for Visual Tracking
C Sminchisescu, A Kanaujia… - IEEE Transactions on …, 2007 - ieeexplore.ieee.org
We introduce BM³E, a Conditional Bayesian Mixture of Experts Markov Model, for consistent
probabilistic estimates in discriminative visual tracking. The model applies to problems of …
probabilistic estimates in discriminative visual tracking. The model applies to problems of …
Structured output-associative regression
L Bo, C Sminchisescu - 2009 IEEE Conference on Computer …, 2009 - ieeexplore.ieee.org
Structured outputs such as multidimensional vectors or graphs are frequently encountered in
real world pattern recognition applications such as computer vision, natural language …
real world pattern recognition applications such as computer vision, natural language …
[图书][B] Intelligent video surveillance: systems and technology
Y Ma, G Qian - 2009 - taylorfrancis.com
From the streets of London to subway stations in New York City, hundreds of thousands of
surveillance cameras ubiquitously collect hundreds of thousands of videos, often running …
surveillance cameras ubiquitously collect hundreds of thousands of videos, often running …
Impact of dynamics on subspace embedding and tracking of sequences
K Moon, V Pavlovic - … Vision and Pattern Recognition (CVPR'06 …, 2006 - ieeexplore.ieee.org
In this paper we study the role of dynamics in dimensionality reduction problems applied to
sequences. We propose a new family of marginal auto-regressive (MAR) models that …
sequences. We propose a new family of marginal auto-regressive (MAR) models that …
[PDF][PDF] Monocular 3D tracking of articulated human motion in silhouette and pose manifolds
F Guo, G Qian - EURASIP journal on Image and Video Processing, 2008 - Springer
This paper presents a robust computational framework for monocular 3D tracking of human
movement. The main innovation of the proposed framework is to explore the underlying data …
movement. The main innovation of the proposed framework is to explore the underlying data …
3D human motion tracking using manifold learning
F Guo, G Qian - 2007 IEEE International Conference on Image …, 2007 - ieeexplore.ieee.org
This paper introduces a framework to track 3D human movement using Gaussian process
dynamic model (GPDM) and particle filter. The framework combines the particle filter and …
dynamic model (GPDM) and particle filter. The framework combines the particle filter and …
Feature-based pose estimation
In this chapter we review challenges and methodology for feature-based predictive three-
dimensional human pose reconstruction, based on image and video data. We argue that …
dimensional human pose reconstruction, based on image and video data. We argue that …