Discovering discriminative action parts from mid-level video representations

By Michalis Raptis, Iasonas Kokkinos, Stefano Soatto. 2012

249 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

Delving into Egocentric Actions (2015, neutral): “Raptis et al. [28] extract action parts by forming clusters of trajectories.”
Finding action tubes (2014, methodology): “[32] extract clusters of trajectories and group them to predict an action class using a graphical model.”
Temporal Segment Networks for Action Recognition in Videos (2017, background): “mid-level representations have been proposed for action recognition [3], [23], [47], [48], [49], [50], [51].”
A Hierarchical Pose-Based Approach to Complex Action Understanding Using Dictionaries of Actionlets and Motion Poselets (2016, methodology): “Inspired by [23], we use the Cattell’s Scree test to find a suitable number of actionlets to model each atomic action.”
Action Classification with Locality-Constrained Linear Coding (2014, methodology): “Note that [20]–[22] used the raw videos whereas our method uses the ROI of the videos.”
A Review of Human Activity Recognition Methods (2015, methodology): “Raptis et al. (2012) proposed a mid-level approach extracting spatiotemporal features and constructing clusters of trajectories, which could be considered as candidates of an action.”
Moving Poselets: A Discriminative and Interpretable Skeletal Motion Representation for Action Recognition (2015, methodology): “An important disadvantage of all these methods for building mid-level representations for action classification is that, except for [19, 4, 14], the mid-level codebook/classifier is learned separately from the action classifiers using clustering/mining techniques, which might not be discriminative f”
Video Event Understanding Using Natural Language Descriptions (2013, background): “Weakly supervised action models Discriminative spatio temporal regions in videos or images to localize the actions in [27, 21, 28].”
Rank Pooling for Action Recognition (2015, background): “Classifiers trained on this new representation turn out to be remarkably good at distinguishing actions.”
Learning latent spatio-temporal compositional model for human action recognition (2013, methodology): “[26] extracted clusters of trajectories and proposed a graphical model to incorporate constraints for individual and group events.”
Spatiotemporal Deformable Part Models for Action Detection (2013, methodology): “Finally, we focus explicitly on demonstrating the effectiveness of action detection within a DPM framework, without resorting to global bag-of-words information [12, 17], trajectories [17] or expensive video segmentation [2, 9].”
Action Recognition with Actons (2013, background): “, discriminative parts/patches [22, 23, 29, 28] or semantical attributes [25, 15, 14]) for image/video recognition has been a recent active area in computer vision.”
Activity representation with motion hierarchies (2013, unknown): “We also differ from global hierarchies of local features [Mikolajczyk and Uemura, 2008, Reddy et al, 2009] and shape-motion prototypes [Jiang et al, 2012b].”
Listen to Look: Action Recognition by Previewing Audio (2019, methodology): “Research has transitioned from initial methods using hand-crafted local spatiotemporal features [35, 66, 61] to mid-level descriptors [46, 29, 62], and more recently to deep video representations learned end-to-end [50, 30, 16].”
Blocks That Shout: Distinctive Parts for Scene Classification (2013, background): “[25] action parts capture spatio-temporal components of human activities.”
Motionlets: Mid-level 3D Parts for Human Motion Recognition (2013, background): “In [25], Raptis et al. group the trajectories into clusters, each of which can be seen as an action part.”
Action Recognition by Hierarchical Mid-Level Action Elements (2015, methodology): “Among all of the methods, [32], [24], [11] and our full model use hierarchical structures.”
From Actemes to Action: A Strongly-Supervised Representation for Detailed Action Understanding (2013, background): “Here, parts have been proposed that capture a neighborhood of spacetime [13, 7, 29, 22], a spatial keyframe [26] or a temporal segment while considering the entire spatial extent of the scene [21, 9].”