Efficient Nonlinear Markov Models for Human Motion

Lehrmann, Andreas M.; Gehler, Peter V.; Nowozin, Sebastian

doi:10.1109/cvpr.2014.171

Cited by 152 publications

(96 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The size of the latent space is the number of frames per video sequence, that is on average 1200 frames per sequence. The score of a specific frame in a video sequence x is w T φ(x, y, h) = w T y φ(x, h) where w y are the parameters that correspond to the label y and φ(x, h) is the feature vector extracted from the video sequence frame h. We take the same features derived from 3D joint locations as in [11,14,20], obtaining a feature vector φ(x, h) of dimension 130. The loss Δ is the standard 0-1 loss.…”

Section: Multi-class Gesture Recognitionmentioning

confidence: 99%

See 1 more Smart Citation

Entropy-Based Latent Structured Output Prediction

Bouchacourt

Nowozin²,

Kumar

2015

2015 IEEE International Conference on Computer Vision (ICCV)

Self Cite

View full text Add to dashboard Cite

Recently several generalizations of the popular latent structural SVM framework have been proposed in the literature. Broadly speaking, the generalizations can be divided into two categories: (i) those that predict the output variables while either marginalizing the latent variables or estimating their most likely values; and (ii) those that predict the output variables by minimizing an entropy-based uncertainty measure over the latent space. In order to aid their application in computer vision, we study these generalizations with the aim of identifying their strengths and weaknesses. To this end, we propose a novel prediction criterion that includes as special cases all previous prediction criteria that have been used in the literature. Specifically, our framework's prediction criterion minimizes the Aczél and Daròczy entropy of the output. This in turn allows us to design a learning objective that provides a unified framework (UF) for latent structured prediction. We develop a single optimization algorithm and empirically show that it is as effective as the more complex approaches that have been previously employed for latent structured prediction. Using this algorithm, we provide empirical evidence that lends support to prediction via the minimization of the latent space uncertainty.

show abstract

Section: Multi-class Gesture Recognitionmentioning

confidence: 99%

“…The original purpose of the data set was to enable research into low latency gesture detection [20]. However, it has since been used to classify the sequence as a whole [11,14]. We also perform sequence classification but do not use the individual frame-level annotations.…”

Section: Multi-class Gesture Recognitionmentioning

confidence: 99%

Entropy-Based Latent Structured Output Prediction

Bouchacourt

Nowozin²,

Kumar

2015

2015 IEEE International Conference on Computer Vision (ICCV)

Self Cite

View full text Add to dashboard Cite

show abstract

“…Euler angles are commonly used in biomechanics and kinematic analysis [27,42,58,65,66], because they are very intuitive and easy to interpret rotation description method. In case we are working on time varying signals averaging we have to deal with two factors: signals might vary in length and periodicity.…”

Section: Data Preprocessingmentioning

confidence: 99%

“…The human motion's models based on training on MoCap data typically use Markov Models [42], graph representation [13] or Dynamic Time Warping (DTW) [1,51,61].Very often those methods do not employ full body evaluation or operate in reduced PCA space [14] in which not all features are taken into account during calculation.…”

Section: Introductionmentioning

confidence: 99%

Averaging of motion capture recordings for movements’ templates generation

Hachaj

Koptyra

Ogiela

2018

Multimed Tools Appl

View full text Add to dashboard Cite

In this paper we propose, describe and evaluate the novel motion capture (MoCap) data averaging framework. It incorporates hierarchical kinematic model, angle coordinates' preprocessing methods, that recalculate the original MoCap recording making it applicable for further averaging algorithms, and finally signals averaging processing. We have tested two signal averaging methods namely Kalman Filter (KF) and Dynamic Time Warping barycenter averaging (DBA). The propose methods have been tested on MoCap recordings of elite Karate athlete, multiple champion of Oyama karate knockdown kumite who performed 28 different karate techniques repeated 10 times each. The proposed methods proved to have not only high effectiveness measured with root-mean-square deviation (4.04 ± 5.03 degrees for KF and 5.57 ± 6.27 for DBA) and normalized Dynamic Time Warping distance (0.90 ± 1.58 degrees for KF and 0.93 ± 1.23 for DBA), but also the reconstruction and visualization of those recordings persists all crucial aspects of those complicated actions. The proposed methodology has many important applications in classification, clustering, kinematic analysis and coaching. Our approach generates an averaged full body motion template that can be practically used for example for human actions Multimed Tools Appl (2018) 77:30353-30380 recognition. In order to prove it we have evaluated templates generated by our method in human action classification tasks using DTW classifier. We have made two experiments. In first leave -one -out cross -validation we have obtained 100% correct recognitions. In second experiment when we classified recordings of one person using templates of another recognition rate 94.2% was obtained.

show abstract

“…Hidden Markov Models (HMMs) and their extensions are widely used for classification of activities in many applications, including vision-based systems, such as [7]. Evidence is usually encoded as observable random variables and activities are represented as the hidden states of a Markov process, which models the behavior of the observed agent.…”

Section: Introductionmentioning

confidence: 99%

Intent Understanding Using an Activation Spreading Architecture

2015

View full text Add to dashboard Cite

Abstract:In this paper, we propose a new approach for recognizing intentions of humans by observing their activities with a color plus depth (RGB-D) camera. Activities and goals are modeled as a distributed network of inter-connected nodes in an Activation Spreading Network (ASN). Inspired by a formalism in hierarchical task networks, the structure of the network captures the hierarchical relationship between high-level goals and low-level activities that realize these goals. Our approach can detect intentions before they are realized and it can work in real-time. We also extend the formalism of ASNs to incorporate contextual information into intent recognition. We further augment the ASN formalism with special nodes and synaptic connections to model ordering constraints between actions, in order to represent and handle partial-order plans in our ASN. A fully functioning system is developed for experimental evaluation. We implemented a robotic system that uses our intent recognition to naturally interact with the user. Our ASN based intent recognizer is tested against three different scenarios involving everyday activities performed by a subject, and our results show that the proposed approach is able to detect low-level activities and recognize high-level intentions effectively in real-time. Further analysis shows that contextual and partial-order ASNs are able to discriminate between otherwise ambiguous goals.

show abstract

Efficient Nonlinear Markov Models for Human Motion

Cited by 152 publications

References 27 publications

Entropy-Based Latent Structured Output Prediction

Entropy-Based Latent Structured Output Prediction

Averaging of motion capture recordings for movements’ templates generation

Intent Understanding Using an Activation Spreading Architecture

Contact Info

Product

Resources

About