A 3-dimensional sift descriptor and its application to action recognition

By Paul Scovanner, Saad Ali, Mubarak Shah. 2007

1,827 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

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

In Computer Vision and Pattern Recognition

Citations in the corpus

Differential Recurrent Neural Networks for Action Recognition (2015, methodology): “For example, HOG3D [15] uses the histogram of 3D gradient orientations to represent the motion structure over the frame sequences; 3D-SIFT [26] extends the popular SIFT descriptor to characterize the scale-invariant spatio-temporal structure f or 3D video volume; actionlet ensemble [ 31] utilizes a ”
C3D: Generic Features for Video Analysis (2014, positive): “SIFT and HOG are also extended into 3 D into SIFT-3 D [41] and HOG3D [25] for action recognition.”
A Closer Look at Spatiotemporal Convolutions for Action Recognition (2017, background): “Some proposed video representations include spatiotemporal interest points (STIPs) [21], SIFT-3D [27], HOG3D [18], Motion Boundary Histogram [5], Cuboids [6], and ActionBank [26].”
Video In Sentences Out (2012, methodology): “…means that one cannot use common approaches to event recognition, such as spatiotemporal bags of words (Laptev et al., 2007; Niebles et al., 2008; Scovanner et al., 2007), spatiotemporal volumes (Blank et al., 2005; Laptev et al., 2008; Rodriguez et al., 2008), and tracked feature points (Liu et…”
Spatiotemporal Contrastive Video Representation Learning (2020, background): “area have carried over successfully from the static image domain to the video domain. Classic hand-crafted local invariant features (e.g., SIFT [43]) for images have their counterparts (e.g., 3D SIFT [54]) in videos, where the temporal dimension of videos gives rise * The ﬁrst two authors contribute”
Action recognition by dense trajectories (2011, background): “Experimental results show a significant improvement over the state of the art on four datasets of varying difficulty, i.e. KTH, YouTube, Hollywood2 and UCF sports.”
Spatio-temporal relationship match: Video structure comparison for recognition of complex human activities (2009, methodology): “[3, 14]) detects each interest point locating a salient change in a video volume, and constructs a descriptor summarizing a small video patch around it.”
Human activity analysis (2011, background): “These approaches are particularly motivated by the success of the object recognition methodologies using sparse local appearance features, such as SIFT descriptors [Lowe 1999].”
Structure-Preserving Binary Representations for RGB-D Action Recognition (2016, methodology): “As 3D versions of SURF [25], SIFT [8] and HOF [11], 3D speeded up robust features (SURF3D) [26], 3D scale invariant feature transforms (3D-SIFT) [27] and 3D motion features [28], [29] have been proposed for action recognition respectively.”
Discriminative subvolume search for efficient action detection (2009, methodology): “Contemporary methods for action categorization use local spatio-temporal features to characterize the video and perform classification over the set of local features [13] [12] [17] [19] To improve the classification performance, both shape and motion information are applied for action categorization”
Optical Flow Guided Feature: A Fast and Robust Motion Representation for Video Action Recognition (2017, methodology): “Traditional methods extracted hand-craft local visual features such as 3DHOG [19], Motion Boundary Histograms (MBH) [8], improved Dense Trajectory (iDT) [40, 39] and then encoded them into sparse or compact feature vectors which were fed into classifiers [27, 26].”
A Review on Human Activity Recognition Using Vision-Based Method (2017, methodology): “Scovanner et al. [46] introduced a 3D SIFT descriptor and used it in HAR.”
Efﬁcient feature extraction, encoding and classiﬁcation for action recognition (neutral): “Several alternative local descriptors have been proposed including histograms of ﬂow orientations (HOF) [16], histograms of 3D gradients (HOG3D) [12, 33], motion boundary histograms (MBH) [4, 35], shapes of point trajectories [20, 26, 35], local trinary patterns [13, 37] and others.”
ASIFT: A New Framework for Fully Affine Invariant Image Comparison (2009, methodology): “The SIFT method and its variants have been popularly applied for scene recognition [10, 40, 50, 61, 15, 52, 65, 41] and detection [14, 46], robot localization [6, 53, 47, 43], image registration [64], image retrieval [16], motion tracking [59, 20], 3D modeling and reconstruction [49, 62], building p”