Free viewpoint action recognition using motion history volumes

By Daniel Weinland, Rémi Ronfard, Edmond Boyer. 2006

1,002 citations recorded by the source corpus.

Subfields: Biomedical Engineering, Computer Vision and Pattern Recognition, Human-Computer Interaction

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biomedical Engineering

In Computer Vision and Pattern Recognition

In Human-Computer Interaction

Citations in the corpus

A survey on data fusion in internet of things: Towards secure and privacy-preserving fusion (2019, methodology): “N N N [63] Activity recognition Camera data Weighted feature fusion Weizmann Dataset[72], MuHAVi Dataset[73], IXMAS Dataset[74] N Y 95.2 - - - - - - - -”
Histogram of Oriented Principal Components for Cross-View Action Recognition (2014, background): “Since HOPC and STK-Distribution capture complimentary information, their combination significantly improves classification accuracy which is over 22% higher than the nearest competitor on both datasets.”
HOPC: Histogram of Oriented Principal Components of 3D Pointclouds for Action Recognition (2014, methodology): “In RGB videos, in order to recognize actions across viewpoint changes, mostly view independent representations are proposed such as view invariant spatio-temporal features [2,20,22,23,25,33].”
A New Representation of Skeleton Sequences for 3D Action Recognition (2017, methodology): “The cylindrical coordinates have been used to extract viewinvariant motion features for action recognition in [47].”
Event Detection in Crowded Videos (2007, background): “These can be broadly categorized into approaches based on tracking [21, 26], flow [8,14,24], spatio-temporal shapes [2,3,31,32], and interest points [7, 20, 23].”
Retrieving actions in movies (2007, background): “Whereas for object recognition there now exist comprehensive datasets with thousands of realistic images for tens or hundreds object classes (Caltech 256, PASCAL VOC2006), the situation is different in action recognition where the existing datasets [16, 24] provide only a few action classes recorded”
3D Pose from Motion for Cross-View Action Recognition via Non-linear Circulant Temporal Encoding (2014, positive): “We chose the INRIA IXMAS dataset [31] for our experiments.”
Human Action Recognition From Various Data Modalities: A Review (2020, neutral): “Dataset Year Modality #Class #Subject #Sample #View KTH [347] 2004 RGB 6 25 2,391 1 Weizmann [40] 2005 RGB 10 9 90 1 IXMAS [348] 2006 RGB 11 10 330 5 HDM05 [349] 2007 RGB,S 130 5 2,337 1 Hollywood [350] 2008 RGB 8 - 430 Hollywood2 [351] 2009 RGB 12 - 3,669 MSR-Action3D [217] 2010 S,D 20 10 567 1 Oly”
Machine Recognition of Human Activities: A Survey (2008, background): “Administration (TSA) airport tarmac data set [9], Free Viewpoint National Institute for Research in Computer Science and Control (INRIA) data set [138], and the Royal Institute of Technology (KTH) actions data set [47].”
Learning a Deep Model for Human Action Recognition from Novel Viewpoints (2016, methodology): “Per class recognition accuracy of our proposed R-NKTM and NKTM [43] on the IXMAS [36] dataset.”
Fuzzy human motion analysis: A review (2014, methodology): “In general, sil houette is the binary representation of the segmented regio ns from the background subtraction techniques, where in HMA, h uman silhouette has proved its sufficiency to describe the activities captured by the video [44, 45, 46].”
Cross-View Action Recognition from Temporal Self-similarities (2008, methodology): “[7, 12] perform full 3D reconstruction from extracting silhouettes seen from multiple deployed cameras.”
Hierarchical Clustering Multi-Task Learning for Joint Human Action Grouping and Recognition (2017, background): “The dimensions of BoWs for MV-TJU (RGB/ depth) and IXMAS were set as 1,000.”
3D Activity Recognition Using Motion History and Binary Shape Templates (2014, neutral): “[23] suggest the use of motion history volumes for analyzing voxelset sequences.”
View-invariant action recognition using fundamental ratios (2008, background): “There are mainly two lines of research to tackle viewinvariance: One is based on using multiple cameras, such as [16, 1, 9, 4], and the second is based on multiple frames of a stationary camera.”
Learning to Recognize Activities from the Wrong View Point (2008, methodology): “Alternatively, one might use an implicit representation of 3D shapes as a set of silhouettes from multiple views [2,26], spatio-temporal volumes derived from such a silhouettes [42,47], or spatio-temporal curvature of 2D trajectory [32].”