Multi-View Human Action Recognition Using Wavelet Data Reduction and Multi-Class Classification

Aryanfar, Alihossein; Yaakob, Razali; Halin, Alfian Abdul; Sulaiman, Nasir; Kasmiran, Khairul Azhar; Mohammadpour, Leila

doi:10.1016/j.procs.2015.08.540

Cited by 10 publications

(4 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Considering the advantages of wavelet transform, A. A et al [31] proposed a novel method for MVHAR by integrating the wavelet transform with silhouette. Initially, the contour of human silhouette is extracted and a distance signal is measured.…”

Section: Related Workmentioning

confidence: 99%

View-Invariant Feature Representation for Action Recognition under Multiple Views

Reddy¹,

Naidu²,

Vardhan³

2019

IJIES

View full text Add to dashboard Cite

Robust Action Recognition under multiple views has gained a significant research interest recently. To enhance the performance of Multi-view Action Recognition, we propose a novel Feature extraction and Feature Selection mechanism that allows building a mutual relationship between the actions sequences of multiple views. The feature extraction considered multiple features which are invariant to scale and orientation. Three different features such as Intensity Features, Orientational Features and Contour Features are used to represent every action. Further, the feature selection is accomplished through self-similarity matrix and is very much helpful in the provision of a perfect discrimination between actions sequences of different views. The proposed method is validated over the standard multi-view IXMAS dataset and experimental results confirm that the proposed method outperforms the conventional approaches with respect to Recognition Accuracy.

show abstract

Section: Related Workmentioning

confidence: 99%

View-Invariant Feature Representation for Action Recognition under Multiple Views

Reddy¹,

Naidu²,

Vardhan³

2019

IJIES

View full text Add to dashboard Cite

show abstract

“…Since our method takes humancentered subvolumes recorded from multiple views as input, , , } ∈cub( ),V=1: , =1: ; Predefined parameters dep max , ∈ (0, 1), ∈ (0, 1), and ∈ (0, 1); Output: Decision tree Tree , ; (1) Build a bootstrap dataset , by random sampling from with replacement; (2) Create a root node and set its depth to 1, then assign all cuboids in , to it; (3) Initialize an unsettled node queue Υ = 0 and push the root node into Υ; (4) while Υ ̸ = 0 do (5) Pop the first node in Υ; (6) if depth of is larger than dep max or cuboids assigned to belong to the same action and position then (7) Label node as a leaf, and then calculate P and Q from cuboids at node ; (8) Add a triple ( , , ) into decision tree Tree , ; (9) else (10) Initialize the feature candidate set Δ = 0; (11) if random number < then (12) Add a set of randomly selected optical flow features to Δ; (13) else (14) Add a set of randomly selected HOG3D features to Δ; (15) end if (16) if random number < then (17) Add two-dimensional temporal context features to Δ; (18) end if (19) maxgain = −∞, generate a random number ; (20) for each ∈ Δ do (21) if < then (22) Search for the corresponding threshold and compute information gain ( ) in terms of action labels of cuboids arriving at ; (23) else (24) Search for the corresponding threshold and compute information gain ( ) in terms of positions of cuboids arriving at ; (25) end if (26) if ( ) > maxgain then (27) * = , * = ; (28) end if (29) end for (30) Create left children node and right children node , set their depth to dep + 1, and assign each cuboid arriving at to or according to * and * ; then push node * and * into Υ; (31) Add a quintuple ( , , , * , * ) into decision tree Tree , ; (32) end if (33) end while (34) return Decision tree Tree , ; Algorithm 1: Construction of a decision tree.…”

Section: Experimental Settingmentioning

confidence: 99%

Learning a Mid-Level Representation for Multiview Action Recognition

Liu

Shi

et al. 2018

Advances in Multimedia

View full text Add to dashboard Cite

Recognizing human actions in videos is an active topic with broad commercial potentials. Most of the existing action recognition methods are supposed to have the same camera view during both training and testing. And thus performances of these single-view approaches may be severely influenced by the camera movement and variation of viewpoints. In this paper, we address the above problem by utilizing videos simultaneously recorded from multiple views. To this end, we propose a learning framework based on multitask random forest to exploit a discriminative mid-level representation for videos from multiple cameras. In the first step, subvolumes of continuous human-centered figures are extracted from original videos. In the next step, spatiotemporal cuboids sampled from these subvolumes are characterized by multiple low-level descriptors. Then a set of multitask random forests are built upon multiview cuboids sampled at adjacent positions and construct an integrated mid-level representation for multiview subvolumes of one action. Finally, a random forest classifier is employed to predict the action category in terms of the learned representation. Experiments conducted on the multiview IXMAS action dataset illustrate that the proposed method can effectively recognize human actions depicted in multiview videos.

show abstract

“…Although, the study is not directly an application of video analytics but its mechanism is highly assistive in video analytics. Aryanfar et al [38] have used a conventional classifier (support vector machine and naïve Bayes) in order to recognize the human action. The outcome of the study was compared with existing mechanism to find approximately 90% of recognition rate.…”

Section: Existing Techniques Of Vamentioning

confidence: 99%

“…Not tested for analytics Aryanfar et al [38] support vector machine and naïve Bayes -Good Recognition rate Not tested for analytics Ayed et al [39] Mining with MapReduce, Wavelets -Reduce processing time -No comparative analysis -No complexity Analysis Cai et al [40] Statistical Analysis Good precision -No comparative analysis -No complexity Analysis Chen et al [41] PeakVizor, Glyph visualization -Interactive detection of user's online behaviour -Doesn't work online -cannot perform classification -No complexity Analysis Kim et al [42] Morphological study of facial expression -ideal for online games -No comparative analysis -No complexity Analysis Mao et al [43] Video traceability Accurate target extraction -Doesn't include high level of data extraction or mining.…”

Section: Existing Techniques Of Vamentioning

confidence: 99%

Insights to Video Analytic Modelling Approach with Future Line of Research

Chandra¹,

Reddy²

2016

IJCA

View full text Add to dashboard Cite

The area of Video analytics has made some significant improvement due to advancement in image processing and datamining techniques. However, the inclination is still towards image contents and less to mined contents owing to many unsolved issues. Although concept of mining is more than 2 decade old, but mining approaches are yet to be standardized in the area of video surveillance system. With evolution of newer set of challenges in video capturing, existing mining models finds itself less applicable due to unstructured format of dynamic frames. Hence, this paper discusses about video analytics and presents a brief discussion of frequently used mining approaches in video as well as discussed some recent studies in this direction in order to scale the degree of effectiveness in existing system. The paper also presents research gap and provided solution as future line of research as a possible way to overcome the research gap.

show abstract

Multi-View Human Action Recognition Using Wavelet Data Reduction and Multi-Class Classification

Cited by 10 publications

References 37 publications

View-Invariant Feature Representation for Action Recognition under Multiple Views

View-Invariant Feature Representation for Action Recognition under Multiple Views

Learning a Mid-Level Representation for Multiview Action Recognition

Insights to Video Analytic Modelling Approach with Future Line of Research

Contact Info

Product

Resources

About