Spela Ivekovic

Markerless human articulated tracking using hierarchical particle swarm optimisation

In this paper, we address markerless full-body articulated human motion tracking from multi-view video sequences acquired in a studio environment. The tracking is formulated as a multi-dimensional non-linear optimisation and solved using particle swarm optimisation (PSO), a swarm-intelligence algorithm which has gained popularity in recent years due to its ability to solve difficult non-linear optimisation problems. We show that a small number of particles achieves accuracy levels comparable with several recent algorithms. PSO initialises automatically, does not need a sequence-specific motion model and recovers from temporary tracking divergence through the use of a powerful hierarchical search algorithm (HPSO). We compare experimentally HPSO with particle filter (PF), annealed particle filter (APF) and partitioned sampling annealed particle filter (PSAPF) using the computational framework provided by Balan et al. HPSO accuracy and consistency are better than PF and compare favourably with those of APF and PSAPF, outperforming it in sequences with sudden and fast motion. We also report an extensive experimental study of HPSO over ranges of values of its parameters.

Human Body Pose Estimation with PSO

In this paper we describe the application of Particle Swarm Optimisation to the problem of human body pose estimation from multiple view video sequences. We use a subdivision body model with an underlying skeleton layer to estimate and illustrate the body pose. The optimisation looks for the best match between the silhouettes extracted from the original video sequence and the silhouettes generated by the projection of the model in a pose suggested by the PSO. The original PSO algorithm is applied hierarchically and combined with the full overall optimisation to decrease the effects of error propagation. Results demonstrate the ability of PSO to reliably recover the correct body pose from 4-viewpoint video sequences.

Human body pose estimation with particle swarm optimisation

In this paper we address the problem of human body pose estimation from still images. A multi-view set of images of a person sitting at a table is acquired and the pose estimated. Reliable and efficient pose estimation from still images represents an important part of more complex algorithms, such as tracking human body pose in a video sequence, where it can be used to automatically initialise the tracker on the first frame. The quality of the initialisation influences the performance of the tracker in the subsequent frames. We formulate the body pose estimation as an analysis-by-synthesis optimisation algorithm, where a generic 3D human body model is used to illustrate the pose and the silhouettes extracted from the images are used as constraints. A simple test with gradient descent optimisation run from randomly selected initial positions in the search space shows that a more powerful optimisation method is required. We investigate the suitability of the Particle Swarm Optimisation (PSO) for solving this problem and compare its performance with an equivalent algorithm using Simulated Annealing (SA). Our tests show that the PSO outperforms the SA in terms of accuracy and consistency of the results, as well as speed of convergence.

Dense wide-baseline disparities from conventional stereo for immersive videoconferencing

We propose an algorithm creating consistent, dense disparity maps from incomplete disparity data generated by a conventional stereo system used in a wide-baseline configuration. The reference application is IBR-oriented immersive videoconferencing, in which disparities are used by a view synthesis module to create instantaneous views of remote speakers consistent with the local speakers viewpoint. We perform spline-based disparity interpolation within nonoverlapping regions are defined by discontinuity boundaries identified in the incomplete disparity map. We demonstrate very good results on significantly incomplete disparity data computed by a conventional correlation-based stereo algorithm on a real wide-baseline stereo pair acquired by an immersive videoconferencing system.

Markerless multi-view articulated pose estimation using adaptive hierarchical particle swarm optimisation

In this paper, we present a new adaptive approach to multi-view markerless articulated human body pose estimation from multi-view video sequences, using Particle Swarm Optimisation (PSO). We address the computational complexity of the recently developed hierarchical PSO (HPSO) approach, which successfully estimated a wide range of different motion with a fixed set of parameters, but incurred an unnecessary overhead in computational complexity. Our adaptive approach, called APSO, preserves the black-box property of the HPSO in that it requires no parameter value input from the user. Instead, it adaptively changes the value of the search parameters online, depending on the quality of the pose estimate in the preceding frame of the sequence. We experimentally compare our adaptive approach with HPSO on four different video sequences and show that the computational complexity can be reduced without sacrificing accuracy and without requiring any user input or prior knowledge about the estimated motion type.

Fitting subdivision surface models to noisy and incomplete 3-D data

We describe an algorithm for fitting a Catmull-Clark subdivision surface model to an unstructured, incomplete and noisy data set. We complete the large missing data regions with the a-priori shape information and produce a smooth, compact and structured data description. The result can be used for further data manipulation, compression, or visualisation. Our fitting algorithm uses a quasi-interpolation technique which manipulates the base mesh of the subdivision model to achieve better approximation. We extend the approach designed for scientific visualisation and animation to deal with incomplete and noisy data and preserve prior shape constraints where data is missing. We illustrate the algorithm on range and stereo data with a set of different subdivision models and demonstrate the applicability of the method to the problem of novel view synthesis from incomplete stereo data.

Markerless Human Motion Capture Using Hierarchical Particle Swarm Optimisation

In this paper, we address full-body articulated human motion tracking from multi-view video sequences acquired in a studio environment. The tracking is formulated as a multi-dimensional nonlinear optimisation and solved using particle swarm optimisation (PSO), a swarm-intelligence algorithm which has gained popularity in recent years due to its ability to solve difficult nonlinear optimisation problems. Our tracking approach is designed to address the limits of particle filtering approaches: it initialises automatically, removes the need for a sequence-specific motion model and recovers from temporary tracking divergence through the use of a powerful hierarchical search algorithm (HPSO). We quantitatively compare the performance of HPSO with that of the particle filter (PF), annealed particle filter (APF) and partitioned sampling annealed particle filter (PSAPF). Our test results, obtained using the framework proposed by Balan et al [1] to compare articulated body tracking algorithms, show that HPSO’s pose estimation accuracy and consistency is better than PF, APF and PSAPF.

Robust correspondenceless 3-d iris location for immersive environments

We present a system locating the contour of an iris in space using robust active ellipse search and correspondenceless stereo. Robust iris location is the basis for gaze estimation and tracking, and, as such, an essential module for augmented and virtual reality environments. The system implements a robust active ellipse search based on a multi-scale contour detection model. The search is carried out by a simulated annealing algorithm, guaranteeing excellent performance in spite of heavy occlusions due to blinking, uncontrolled lighting, erratic target motion, and reflections of unpredictable scene elements. Stereo correspondence is avoided altogether by intersecting conjugate epipolar lines with the located ellipses. Experiments on synthetic and real images indicate very good performance of both location and reconstruction modules.

Multiresolution approach to biomedical image segmentation with statistical models of appearance

Structural variability present in biomedical images is known to aggravate the segmentation process. Statistical models of appearance proved successful in exploiting the structural variability information in the learning set to segment a previously unseen medical image more reliably. In this paper we show that biomedical image segmentation with statistical models of appearance can be improved in terms of accuracy and efficiency by a multiresolution approach. We outline two different multiresolution approaches. The first demonstrates a straightforward extension of the original statistical model and uses a pyramid of statistical models to segment the input image on various resolution levels. The second applies the idea of direct coefficient propagation through the Gaussian image pyramid and uses only one statistical model to perform the multiresolution segmentation in a much simpler manner. Experimental results illustrate the scale of improvement achieved by using the multiresolution approaches described. Possible further improvements are discussed at the end.