Towards Understanding Action Recognition

By Hueihan Jhuang, Juergen Gall, Silvia Zuffi, Cordelia Schmid, Michael J. Black. 2013

968 citations recorded by the source corpus.

Subfields: Artificial Intelligence, 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 Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

2D/3D Pose Estimation and Action Recognition Using Multitask Deep Learning (2018, background): “Previous approaches have explored classical methods for features extraction [55, 23], where the key idea is to use body joint locations to select visual features in space and time.”
Skeleton-Based Action Recognition Using Spatio-Temporal LSTM Network with Trust Gates (2017, neutral): “Skeleton-based action recognition has been explored in different aspects in recent years [40], [41], [42], [43], [44], [45], [46], [47], [48], [49], [50], [51], [52], [53], [54], [55], [56], [57].”
Chained Multi-stream Networks Exploiting Pose, Motion, and Appearance for Action Classification and Detection (2017, methodology): “We trained the Fast-Net architecture on the J-HMDB [13] and the MPII [1] action recognition datasets.”
Cross Modal Distillation for Supervision Transfer (2015, background): “Note that the representations ψi ∗ Ms and φ L Md may not have the same dimensions.”
Finding action tubes (2014, methodology): “We evaluate our approach on two widely used datasets, namely UCF Sports [33] and J-HMDB [17].”
Video Action Transformer Network (2018, background): “Some recent work has looked at such ﬁne-grained video understanding [8, 17, 21, 37], but has largely been limited to small datasets like UCF-24 [37, 38] or JHMDB [19].”
Moments in Time Dataset: One Million Videos for Event Understanding (2018, background): “JHMDB [24] has human activity categories with joints annotated.”
End-to-End Learning of Action Detection from Frame Glimpses in Videos (2015, neutral): “2 shows our model structure.”
Tracking Emerges by Colorizing Videos (2018, methodology): “It can also track human pose given keypoints annotated in an initial frame [3].”
P-CNN: Pose-Based CNN Features for Action Recognition (2015, methodology): “We extend the work [19] and design a new CNN-based representation for human actions combining positions, appearance and motion of human body parts.”
Towards Weakly-Supervised Action Localization (2016, neutral): “• The J-HMDB dataset [38] contains 928 short videos with 21 actions, including stand up , run and pour .”
Visual Semantic Role Labeling (2015, methodology): “In particular the J-HMDB [29] and UCF Sports dataset [18] are popular test beds for algorithms that study this task [13].”
A Hierarchical Pose-Based Approach to Complex Action Understanding Using Dictionaries of Actionlets and Motion Poselets (2016, methodology): “For this purpose, we use the sub-JHMDB dataset [11], which focuses on videos depicting 12 actions and where most of the actor body is visible in the image frames.”
Learning to Track for Spatio-Temporal Action Localization (2015, neutral): “J-HMDB [16].”
AVA: A Video Dataset of Spatio-Temporally Localized Atomic Visual Actions (2017, methodology): “Performance numbers on popular action recognition datasets such as UCF101 or JHMDB have gone up considerably in recent years, but we believe that this may present an artiﬁcially rosy picture of the state of the art.”
URVOS: Unified Referring Video Object Segmentation Network with a Large-Scale Benchmark (2020, methodology): “However, their method has been validated in the datasets with limited class di-versities, A2D [33] and J-HMDB [9], which only have 8 and 21 predeﬁned action classes, respectively.”
Rethinking Spatiotemporal Feature Learning: Speed-Accuracy Trade-offs in Video Classification (2017, methodology): “We report performance on two widely adopted video action detection datasets: JHDMB [52] and UCF-101-24 [45].”
Human Action Recognition Based on MOCAP Information Using Convolution Neural Networks (2014, methodology): “human actions [3] and the most informative human joints for recognizing an action [4].”
TAP-Vid: A Benchmark for Tracking Any Point in a Video (2022, positive): “Third, we explore several baseline algorithms, and we compare our point tracking dataset to the closest existing point tracking dataset—JHMDB human keypoint tracking [29]—demonstrating that training using our formulation of the problem can increase performance on this far more limited dataset.”