Quasi-periodic action recognition from monocular videos via 3D human models and cyclic HMMs

Thuc, Hoang Le Uyen; Ke, Shian-Ru; Hwang, Jenq‐Neng; Tuấn, Phạm Văn; Châu, Trương Ngọc

doi:10.1109/atc.2012.6404241

Cited by 3 publications

(1 citation statement)

References 12 publications

(10 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Dynamic Time Warping (DTW) and Hidden Markov Model (HMM) were initially proposed for speech recognition [130,131] and then used in HAR applications. Some methods that have been used in HAR applciations are DTW [173], HMMs [140,164], K-nearest neighbors [69,76], SVM [26,128,136,145], Kalman filters [12,21], and more recently, ANNs and deep networks (to be discussed in Chapter 3). More detailed account of classical machine learning classification methods can be found in [14,207].…”

Section: Introductionmentioning

confidence: 99%

Cross-View Action Recognition

2021

Computer Vision

View full text Add to dashboard Cite

Cross View Action Recognition (CVAR) appraises a system's ability to recognise actions from viewpoints that are unfamiliar to the system. The state of the art methods that train on large amounts of training data rely on variation in the training data itself to increase their ability to tackle viewpoints changes. Therefore, these methods not only require a large scale dataset of appropriate classes for the application every time they train, but also correspondingly large amount of computation power for the training process leading to high costs, in terms of time, effort, funds and electrical energy. In this thesis, we propose a methodological pipeline that tackles change in viewpoint, training on small datasets and employing sustainable amounts of resources. Our method uses the optical flow input with a stream of a pre-trained model as-is to obtain a feature. Thereafter, this feature is used to train a custom designed classifier that promotes view-invariant properties. Our method only uses video information as input, in contrast to another set of methods that approach CVAR by using depth or pose input at the expense of increased sensor costs. We present a number of comparative analysis that aided the design of the pipelines, farther assessing the power of each component in the pipeline. The technique can also be adopted to existing, trained classifiers, with minimal fine-tuning, as this work demonstrates by comparing classifiers including shallow classifiers, deep pre-trained classifiers and our proposed classifier trained from scratch. Additionally, we present a set of qualitative results that promote our understanding of the relationship between viewpoints in the feature-space. i Francesca Odone has been a better PhD advisor and mentor to me than anyone can ever hope to have, and I would like to convey my deepest gratitude for her time, effort, guidance and patience and most importantly, for always keeping her door open for her students. I would also like to thank Nicoletta Noceti for her guidance and our collaborations during the PhD study. I would like to thank University of Genova, DIBRIS, and previous and current colleagues for all the help and plenty of constructive conversations throughout the PhD term with special mention to Damiano Malafronte.My family has been remarkably supportive and encouraging. I would like to thank my mother, Sangeeta Goyal, who calls me her strength but in reality she is mine, my sister, Yamini Goyal, to whom I owe, in big part, my sanity and my father, Subhash Goyal. I could not have done this without them.

show abstract

Section: Introductionmentioning

confidence: 99%

Cross-View Action Recognition

2021

Computer Vision

View full text Add to dashboard Cite

show abstract

Vision Based Human Activity Recognition: A Review

Bux

Angelov

Habib

2016

Advances in Intelligent Systems and Computing

View full text Add to dashboard Cite

Human Action Recognition Based on 3D Human Modeling and Cyclic HMMs

Thuc

Hwang

et al. 2014

ETRI J

View full text Add to dashboard Cite

Human action recognition is used in areas such as surveillance, entertainment, and healthcare. This paper proposes a system to recognize both single and continuous human actions from monocular video sequences, based on 3D human modeling and cyclic hidden Markov models (CHMMs). First, for each frame in a monocular video sequence, the 3D coordinates of joints belonging to a human object, through actions of multiple cycles, are extracted using 3D human modeling techniques. The 3D coordinates are then converted into a set of geometrical relational features (GRFs) for dimensionality reduction and discrimination increase. For further dimensionality reduction, k-means clustering is applied to the GRFs to generate clustered feature vectors. These vectors are used to train CHMMs separately for different types of actions, based on the Baum-Welch re-estimation algorithm. For recognition of continuous actions that are concatenated from several distinct types of actions, a designed graphical model is used to systematically concatenate different separately trained CHMMs. The experimental results show the effective performance of our proposed system in both single and continuous action recognition problems. Keywords I. IntroductionHuman action recognition is a growing topic in video analysis and understanding -one of the most popular areas in the community of computer vision -thanks to its applications in surveillance, entertainment, and healthcare. In surveillance, human action recognition can be used in conjunction with video camera footage to help with the recognition and analysis of human actions. In entertainment, human-computer interaction can be helped to appear more natural via human action recognition, which in turn can help increase the entertainment experience. In healthcare, human action recognition can help detect abnormal gaits or assist in a patient's rehabilitation through an analysis of their actions.However, it is challenging to recognize various human actions due to the high number of degrees of freedom associated with the average human body -namely, variations in human poses; variations in the colors of a person's clothing; changes in lighting and illumination; variations in viewpoints; and frequent self-occlusion. Moreover, the use of monocular video sequences further increases the difficulty for human action recognition.Generally, the two main stages in human action recognition are: the feature extraction and representation stage and the classification stage.In the feature extraction and representation stage, the features or characteristics of video frames, such as silhouette, shape, color, and motion, are extracted and represented in a systematic and efficient way. consists of stacking segmented silhouettes (frame by frame) to form a 3D spatial-temporal shape. In a similar way, Ke and others [2] build STVs, for shape-based matching, from image features that are based on the consecutive silhouettes of objects along a time axis, including spatial-temporal region extraction and region matching. Kim and oth...

show abstract

Quasi-periodic action recognition from monocular videos via 3D human models and cyclic HMMs

Cited by 3 publications

References 12 publications

Cross-View Action Recognition

Cross-View Action Recognition

Vision Based Human Activity Recognition: A Review

Human Action Recognition Based on 3D Human Modeling and Cyclic HMMs

Contact Info

Product

Resources

About