Measuring and reducing observational latency when recognizing actions

By Syed Zain Masood, Christopher Ellis, Adarsh Nagaraja, Marshall F. Tappen, J. Laviola, Rahul Sukthankar. 2011

37 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

Space-Time Representation of People Based on 3D Skeletal Data: A Review (2016, background): “[207] Pairwise Joint Displacement & Temporal Location Variations Conc Lowlv Hand X X X”
Modeling Temporal Dynamics and Spatial Configurations of Actions Using Two-Stream Recurrent Neural Networks (2017, neutral): “To get representations of postures, one straightforward feature is the pairwise joint location difference, which can be simply concatenated [32], or casted into 3D cone bins to build a histogram of 3D joints locations [49] for action recognition.”
Two-person interaction detection using body-pose features and multiple instance learning (2012, background): “The entire dataset has a total of 300 interactions approximately.”
Efficient Interaction Recognition through Positive Action Representation (2013, methodology): “For human activity or behavior representation, the method in [16, 20] detected and recognized different activities through body-pose features, hand position features, and motion information, using the Kinect sensor.”
Survey on Emotional Body Gesture Recognition (2018, methodology): “The UCFKinect [150] was collected using Kinect and skeleton estimation from [151].”
Simultaneous Feature and Body-Part Learning for real-time robot awareness of human behaviors (2017, methodology): “A similar skeleton-based representation was implemented by [16], incorporating pairwise joint distances and temporal joint location changes together.”
EigenJoints-based action recognition using Naïve-Bayes-Nearest-Neighbor (2012, result): “This result is also in accordance with the observations in [10] that a 66% reduction in frames only results in a 6.”
Max-Margin Early Event Detectors (2012, background): “Similarly, Ryoo [15] took a passive approach for early recognition of human activities; he developed two variants of the bag-of-words representation to mainly address the computational issues, not timeliness or accuracy, of the detection process.”
3D Gestural Interaction: The State of the Field (2013, background): “Latency can be broken up into computational latency and observational latency [74, 148].”