Visual tracking via geometric particle filtering on the affine group with optimal importance functions

Junghyun, Kwon; Lee, Kyoung Mu; Park, Frank C.

doi:10.1109/cvpr.2009.5206501

Cited by 66 publications

(66 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Object tracking has previously been performed mostly for color/gray-scale image sequences [1,7,8,9,10]. However, depth images pose different challenges to the tracking algorithm than color/gray-scale images.…”

Section: Related Workmentioning

confidence: 99%

“…In this work, we offer a solution to this problem by combining a recent clustering approach based on surface-model fitting [2] with a particle-filter-based affine-motion-estimation approach [7] with some modifications. Because we use a split-and-merge procedure for region growing which automatically adapts to the dynamic range data, predictions from the particle filter can be incorporated in a straightforward manner by refining seeding and, in consequence, the input to the tracker.…”

Section: Introductionmentioning

confidence: 99%

“…Hence, the SSD between the template and the tracked image region cannot be computed in the same way as for color images [7]. In our approach, we first compute the rigid transform between the 3-D range vectors F t (K) and F t=0 (T ) which minimizes the nearest neighbor distance in the least-squares sense [25], and then compute the distance as the l 1 -norm of the difference between the corresponding points, i.e, the measurement function…”

mentioning

confidence: 99%

“…For projective motion estimation, we compute X t using a particle filter, whose state dynamics is based on a constant velocity model as described in [7], i.e.,…”

mentioning

confidence: 99%

See 3 more Smart Citations

Robust surface tracking in range image sequences

Husain

Dellen

Torras

2014

Digital Signal Processing

View full text Add to dashboard Cite

A novel robust method for surface tracking in range-image sequences is presented which combines a clustering method based on surface models with a particle-filter-based 2-D affine-motion estimator. Segmented regions obtained at previous time steps are used to create seed areas by comparing measured depth values with those obtained from surface-model fitting. The seed areas are further refined using a motion-probability region estimated by the particlefilter-based tracker through prediction of future states. This helps resolving ambiguities that arise when surfaces belonging to different objects are in physical contact with each other, for example during hand-object manipulations. Region growing allows recovering the complete segment area. The obtained segmented regions are then used to improve the predictions of the tracker for the next frame. The algorithm runs in quasi real-time and uses on-line learning, eliminating the need to have a priori knowledge about the surface being tracked. We apply the method to in-house depth videos acquired with both time-of-flight and structured-light sensors, demonstrating object tracking in real-world scenarios, and we compare the results with those of an ICP-based tracker.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 99%

“…For projective motion estimation, we compute X t using a particle filter, whose state dynamics is based on a constant velocity model as described in [7], i.e.,…”

mentioning

confidence: 99%

See 2 more Smart Citations

Robust surface tracking in range image sequences

Husain

Dellen

Torras

2014

Digital Signal Processing

View full text Add to dashboard Cite

show abstract

“…Note that the proposed algorithm handles states evolving on the Riemannian manifold rather than the conventional Euclidean space, thus is different from most existing particle filters and their variations. Bayesian recursive filtering using particles has been proposed for specific manifolds in the context of tracking [17][18][19][20], while the following approach is generally applicable for various alignment manifolds. Furthermore, we formulate the static optimization problem into a dynamic state space model, which provides insight on applications of SIS to new problems beyond tracking.…”

Section: Sequential Importance Sampling On the Manifold For Optimal Amentioning

confidence: 99%

Aligning Spatio-Temporal Signals on a Special Manifold

Chellappa

2010

Computer Vision – ECCV 2010

View full text Add to dashboard Cite

Abstract. We investigate the spatio-temporal alignment of videos or features/signals extracted from them. Specifically, we formally define an alignment manifold and formulate the alignment problem as an optimization procedure on this non-linear space by exploiting its intrinsic geometry. We focus our attention on semantically meaningful videos or signals, e.g., those describing or capturing human motion or activities, and propose a new formalism for temporal alignment accounting for executing rate variations among realizations of the same video event. By construction, we address this static and deterministic alignment task in a dynamic and stochastic manner: we regard the search for optimal alignment parameters as a recursive state estimation problem for a particular dynamic system evolving on the alignment manifold. Consequently, a Sequential Importance Sampling iteration on the alignment manifold is designed for effective and efficient alignment. We demonstrate the performance on several types of input data that arise in vision problems.

show abstract

Latest Developments of Gesture Recognition for Human–Robot Collaboration

Liu

Wang

2021

Advanced Human-Robot Collaboration in Manufacturing

View full text Add to dashboard Cite

Large Language Models (LLMs) are gaining popularity in the field of robotics. However, LLM-based robots are limited to simple, repetitive motions due to the poor integration between language models, robots, and the environment. This paper proposes a novel approach to enhance the performance of LLM-based autonomous manipulation through Human-Robot Collaboration (HRC). The approach involves using a prompted GPT-4 language model to decompose high-level language commands into sequences of motions that can be executed by the robot. The system also employs a YOLO-based perception algorithm, providing visual cues to the LLM, which aids in planning feasible motions within the specific environment. Additionally, an HRC method is proposed by combining teleoperation and Dynamic Movement Primitives (DMP), allowing the LLM-based robot to learn from human guidance. Real-world experiments have been conducted using the Toyota Human Support Robot for manipulation tasks. The outcomes indicate that tasks requiring complex trajectory planning and reasoning over environments can be efficiently accomplished through the incorporation of human demonstrations.

show abstract

Visual tracking via geometric particle filtering on the affine group with optimal importance functions

Cited by 66 publications

References 22 publications

Robust surface tracking in range image sequences

Robust surface tracking in range image sequences

Aligning Spatio-Temporal Signals on a Special Manifold

Latest Developments of Gesture Recognition for Human–Robot Collaboration

Contact Info

Product

Resources

About