Particle Swarm Optimization with Soft Search Space Partitioning for Video-Based Markerless Pose Tracking

Fleischmann, Patrick; Austvoll, Ivar; Kwolek, Bogdan

doi:10.1007/978-3-642-33140-4_42

Cited by 7 publications

(22 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Minimal user input is required, in the form of pinpointing the joints on top of the drawing. To avoid forward-backward ambiguity, the proposed method uses a combination of various descriptors [Fleischmann et al 2012]. There are currently no other methods that perform automatic posing for the layout phase of animation that work irrespective of the character model and that can be applied without the need for a pre-existing database.…”

Section: Proposed Approachmentioning

confidence: 99%

Posing 3D Models from Drawings

Gouvatsos

Xiao

Marsden³

et al. 2017

Comput. Entertain.

View full text Add to dashboard Cite

Inferring the 3D pose of a character from a drawing is a complex and under-constrained problem. Solving it may help automate various parts of an animation production pipeline such as pre-visualization. In this article, a novel way of inferring the 3D pose from a monocular 2D sketch is proposed. The proposed method does not make any external assumptions about the model, allowing it to be used on different types of characters. The inference of the 3D pose is formulated as an optimization problem and a parallel variation of the Particle Swarm Optimization algorithm called PARAC-LOAPSO is utilized for searching the minimum. Testing in isolation as well as part of a larger scene, the presented method is evaluated by posing a lamp, a horse, and a human character. The results show that this method is robust, highly scalable, and able to be extended to various types of models.

show abstract

Section: Proposed Approachmentioning

confidence: 99%

Posing 3D Models from Drawings

Gouvatsos

Xiao

Marsden³

et al. 2017

Comput. Entertain.

View full text Add to dashboard Cite

show abstract

“…To initialize optimization in each frame, a motion model is employed to propagate particles to the new frame. Tracking algorithms presented in [2,5,14,15] are examples of algorithms that use zero motion with additional Gaussian noise as motion model: (1) The standard deviation in the Σ matrix is equal to the maximum absolute inter-frame differences of the joint angles that are almost determined in a training process. Experiments on the motion model conducted in [6], show that tracking accuracy is largely dependent on the amount of standard deviation values in the Σ matrix.…”

Section: B Propagation Modelmentioning

confidence: 99%

“…Other five stage are relative to optimization of left and right hand, left and right leg and head orientation, respectively. To avoid error accumulation [2] the standard deviations for first six optimized parameters are reduced to one tenth in these five stages and a trivial search about the result of stage one is performed in the other stages.…”

Section: Likelihood Evaluationmentioning

confidence: 99%

“…3D human motion tracking can be formulated as a nonlinear optimization problem. Gradient based methods are faster than gradient free methods but one of the major problems in these methods is the local optima convergence that yields poor tracking results [2]. Particle filter (PF) can alleviate this problem, but a major challenge is that for successful tracking, the required particles' count increases exponentially with the dimension of the state space [3].…”

Section: Introductionmentioning

confidence: 99%

“…In [7], GA algorithm was proposed to estimate upperbody pose and Zhao and Liu presented an annealed genetic algorithm (AGA) for 3D human motion analysis [8]. Particle Swarm Optimization (PSO) is an evolutionary optimization algorithm that was used in 2006 to solve upper-body pose estimation [9] and has recently got more attention for full-body pose estimation [2,[10][11][12][16][17]. The particles in PSO communicate with each other, therefore, searching is more efficient than the crossover operation in GA [5].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Marker based human pose estimation using annealed particle swarm optimization with search space partitioning

Sharifi

Harati

Vahedian

2014

2014 4th International Conference on Computer and Knowledge Engineering (ICCKE)

View full text Add to dashboard Cite

In this paper, a marker based human pose estimation from multi-view video sequences is presented. The pose estimation problem is defined as optimization of the 45 parameters which define body pose model and is solved using particle swarm optimization (PSO). The objective of this optimization is to maximize a fitness function which formulates how much body model matches with 2D marker's coordinate in video frames. In this algorithm a sampling covariance matrix is used in the first part of the velocity equation of PSO that is annealed with iterations. One of the major problems of this algorithm is the high number of parameters that define the pose of the body model. To tackle this problem, we divide the optimization into six stages that exploit the hierarchical structure of the model. The first stage optimizes the six parameters that define the global orientation and position of the body. Other stages are related to optimization of right and left hand, right and left leg and head orientation, respectively. Experimental results on PEAR 1 dataset [1] indicate that the proposed algorithm can achieve lower estimation error in tracking human motion compared with Annealed Particle Filter (APF) and Standard Particle Filter (PF). Keywords-marker based human motion tracking; particle swarm optimization (PSO)I.

show abstract