An Unsupervised Framework for Action Recognition Using Actemes

Kulkarni, Kaustubh; Boyer, Edmond; Horaud, Radu; Kale, Amit

doi:10.1007/978-3-642-19282-1_47

Cited by 3 publications

(4 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Kaiser ( MoPrim Dataset (Reng et al, 2005) 56 Krüger and Grest (2007) HMM Poses MoPrim Dataset (Reng et al, 2005) 57 Krüger and Herzog (2013) PHMM, Bayes Kulkarni et al (2011) k-means, one-pass DP decoding, HMM 3D Visual Hull IXMAS Dataset (Weinland et al, 2006) 61 Kulkarni et al (1989) Truth Weiz-mann Dataset (Blank et al, 2005) IXMAS Dataset (Weinland et al, 2006) Robust Dataset (Gorelick et al, 2007) 102 Roh et al (2010) VMT, PMT Weiz-mann Dataset (Blank et al, 2005) MSR-II Dataset (Cao et al, 2010) 106 Roshtkhari and Levine 2012Bag of video words, code book STV KTH Dataset (Schldt et al, 2004) Weiz-mann Dataset (Blank et al, 2005) MSR-II Dataset (Cao et al, 2010) 107 (Liu et al, 2009) Hol-lywood2 Dataset (Marszalek et al, 2009) HMDB Dataset (Kuehne et al, 2011) 112 Weizmann Dataset (Blank et al, 2005) UCF-Sports Dataset…”

Section: Resultsmentioning

confidence: 99%

Action representations in robotics: A taxonomy and systematic classification

Zech

Renaudo

Haller

et al. 2019

The International Journal of Robotics Research

View full text Add to dashboard Cite

Understanding and defining the meaning of "action" is substantial for robotics research. This becomes utterly evident when aiming at equipping autonomous robots with robust manipulation skills for action execution. Unfortunately, to this day we still lack both a clear understanding of the concept of an action and a set of established criteria that ultimately characterize an action. In this survey we thus first review existing ideas and theories on the notion and meaning of action. Subsequently we discuss the role of action in robotics and attempt to give a seminal definition of action in accordance with its use in robotics research. Given this definition we then introduce a taxonomy for categorizing action representations in robotics along various dimensions. Finally, we provide a systematic literature survey on action representations in robotics where we categorize relevant literature along our taxonomy. After discussing the current state of the art we conclude with an outlook towards promising research directions.

show abstract

Section: Resultsmentioning

confidence: 99%

Action representations in robotics: A taxonomy and systematic classification

Zech

Renaudo

Haller

et al. 2019

The International Journal of Robotics Research

View full text Add to dashboard Cite

show abstract

“…We will compare with this result with our experiments on this dataset. There are several related works on action recognition with IXMAS dataset, for example [23][24][25]43]. Gu et al [44] listed all state of art experimental results on this dataset.…”

Section: Resultsmentioning

confidence: 99%

“…A challenge of this solution is the inherent ambiguities between 2D image features and 3D poses. Some researchers use multiple-view videos [23][24][25], although single-view image sequences are more generic and easy to acquire. Moreover, recent work shows that even in monocular image sequences, reconstruction ambiguity can be tackled using regression methods like relevance vector machine (RVM) [26].…”

Section: Introductionmentioning

confidence: 99%

Human action recognition based on estimated weak poses

Gong

Gonzàlez

Roca

2012

EURASIP J. Adv. Signal Process.

View full text Add to dashboard Cite

We present a novel method for human action recognition (HAR) based on estimated poses from image sequences. We use 3D human pose data as additional information and propose a compact human pose representation, called a weak pose, in a low-dimensional space while still keeping the most discriminative information for a given pose. With predicted poses from image features, we map the problem from image feature space to pose space, where a Bag of Poses (BOP) model is learned for the final goal of HAR. The BOP model is a modified version of the classical bag of words pipeline by building the vocabulary based on the most representative weak poses for a given action. Compared with the standard k-means clustering, our vocabulary selection criteria is proven to be more efficient and robust against the inherent challenges of action recognition. Moreover, since for action recognition the ordering of the poses is discriminative, the BOP model incorporates temporal information: in essence, groups of consecutive poses are considered together when computing the vocabulary and assignment. We tested our method on two well-known datasets: HumanEva and IXMAS, to demonstrate that weak poses aid to improve action recognition accuracies. The proposed method is scene-independent and is comparable with the state-of-art method.

show abstract

“…They use "strange attractors" to represent the dynamics of time series for action and dynamic texture synthesis, yet do not provide a way to compare two series of observations. Other approaches, inspired from speech and gesture recognition, represent actions as sequences of states [3,13,18,26] or use dynamic probabilistic graphical models [2,21,35,36] to model the temporal aspects of the videos. A limitation of these methods is that they only measure alignments between videos.…”

Section: Introductionmentioning

confidence: 99%

A time series kernel for action recognition

Gaidon

Harchaoui²,

Schmid³

2011

Procedings of the British Machine Vision Conference 2011

View full text Add to dashboard Cite

We address the problem of action recognition by describing actions as time series of frames and introduce a new kernel to compare their dynamical aspects. Action recognition in realistic videos has been successfully addressed using kernel methods like SVMs. Most existing approaches average local features over video volumes and compare the resulting vectors using kernels on bags of features. In contrast, we model actions as time series of per-frame representations and propose a kernel specifically tailored for the purpose of action recognition. Our main contributions are the following: (i) we provide a new principled way to compare the dynamics and temporal structure of actions by computing the distance between their auto-correlations, (ii) we derive a practical formulation to compute this distance in any feature space deriving from a base kernel between frames and (iii) we report experimental results on recent action recognition datasets showing that it provides useful complementary information to the average distribution of frames, as used in state-of-the-art models based on bag-of-features.

show abstract

An Unsupervised Framework for Action Recognition Using Actemes

Cited by 3 publications

References 22 publications

Action representations in robotics: A taxonomy and systematic classification

Action representations in robotics: A taxonomy and systematic classification

Human action recognition based on estimated weak poses

A time series kernel for action recognition

Contact Info

Product

Resources

About