Motion synthesis from annotations

By Okan Arikan, D.A. Forsyth, James F. O'Brien. ACM Transactions on Graphics, 2003

475 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Control and Systems Engineering

View Paper: https://www.semanticscholar.org/paper/79ad3cf586161ba329ce591cd43b96cef0693ba9

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

In Control and Systems Engineering

Citations in the corpus

Segmenting Motion Capture Data into Distinct Behaviors (2004, methodology): “In graphics, Arikan, Forsyth, and O’Brien [3] describe a model-based approach to motion annotation that could also be employed for segmentation.”
Synthesis and evaluation of linear motion transitions (2008, background): “Other researchers [Arikan and Forsyth 2002; Arikan et al. 2003; Kovar and Gleicher 2003] specify transitions by the center of the blend; that is, the speci.ed frames are the midpoint in the blending interval.”
Modeling Human Motion Using Binary Latent Variables (2006, background): “The simplest way to generate new motion sequences based on data is to concatenate parts of training sequences [3].”
Two Distributed-State Models For Generating High-Dimensional Time Series (2011, methodology): “Perhaps the simplest way to generate new motion sequences based on data isto sensibly concatenate short examples from a motion database to meet sparse user-specified constraints (Tanco and Hilton, 2000; Arikan and Forsyth, 2002; Kovar et al., 2002; Lee et al., 2002;Arikan et al., 2003).”
Searching Video for Complex Activities with Finite State Models (2007, methodology): “Motion sequences from this collection are sorted into activities using the labelling of [2].”
Manifold Warp Segmentation of Human Action (2018, background): “[8] point out (claim) that continuity is a signifi-”
Ego-Pose Estimation and Forecasting As Real-Time PD Control (2019, methodology): “Most existing methods only estimate the pose of visible body parts [23, 24, 41, 2, 45].”
Style-based inverse kinematics (2004, background): “…described systems for producing new sequences of movements from examples, either by direct copying and blending of poses [Arikan and Forsyth 2002; Arikan et al. 2003; Kovar et al. 2002; Lee et al. 2002; Pullen and Bregler 2002] or by learning a likelihood function over sequences [Brand and…”
Style translation for human motion (2005, methodology): “Database approaches assemble short subsequences to synthesize motions that match user specifications, such as a desired path on the ground or a list of actions to perform [Kovar et al. 2002a; Lee et al. 2002; Arikan et al. 2003; Arikan and Forsyth 2002].”
QuaterNet: A Quaternion-based Recurrent Model for Human Motion (2018, methodology): “In that context, the task of generating human motion sequences has been addressed with different strategies ranging from concatenative approaches [3] to hidden Markov models [51], switching linear dynamic systems [43], restricted Boltzmann machines [52], Gaussian processes [60], and random forests [”