This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. IEEE TRANSACTIONS ON PATTERN ANALYSIS…

2007

Citations in the corpus

• Manifold Gaussian Processes for regression (2014, background): “This kind of GP training is also analytically intractable, although approximations exist (Lawrence, 2005; Wang et al., 2008; McHutchon and Rasmussen, 2011; Titsias and Lawrence, 2010).”
• Structural-RNN: Deep Learning on Spatio-Temporal Graphs (2015, methodology): “Several approaches based on Gaussian processes [59, 63], Restricted Boltzmann Machines (RBMs) [57, 56, 51], and RNNs [14] have been proposed to model human motion.”
• Trajectron++: Dynamically-Feasible Trajectory Forecasting with Heterogeneous Data (2020, methodology): “Since then, many approaches have been applied to the problem of trajectory forecasting, formulating it as a time-series regression problem and applying methods like Gaussian Process Regression (GPR) [38,47], Inverse Reinforcement Learning (IRL) [32], and Recurrent Neural Networks (RNN) [1,33,46] to ”
• Capturing the human figure through a wall (2015, background): “Prior art has also investigated motion capture in partial occlusions, e.g., [Li et al. 2010; Herda et al. 2000; Liu and McMillan 2006; Park and Hodgins 2006; Chai and Hodgins 2005; Wang et al. 2008].”
• Nonparametric guidance of autoencoder representations using label information (2012, unknown): “In the basic GPLVM, the optimal x n are found by maximizing this marginal likelihood.”
• Recurrent Network Models for Human Dynamics (2015, methodology): “Prior work has tackled synthesis of stylized human motion using bilinear spatiotemporal basis models [1], Hidden Markov Models [3], linear dynamical systems [21], and Gaussian process latent variable models [46, 40], as well as multilinear variants thereof [12, 45].”
• Mode-adaptive neural networks for quadruped motion control (2018, methodology): “Kernel-based approaches [Levine et al. 2012; Wang et al. 2008] can model nonlinearity in motion, but the trained model is limited to express a specific type of motion.”
• Expectation Propagation in Gaussian Process Dynamical Systems (2012, methodology): “For trials 1–7 (403 data points), we used the GPDM [20] to learn MAP estimates of the latent states xt ∈ R(3).”
• Will the Pedestrian Cross? A Study on Pedestrian Path Prediction (2014, methodology): “gory of nonlinear first-order models with augmented visual features, we propose a novel pedestrian path prediction approach, based on Gaussian process dynamical models (GPDMs) [3] and”
• Sharing Features among Dynamical Systems with Beta Processes (2009, background): “Other approac hes develop non-linear dynamical models using Gaussian processes [25] or based on a collection of bin ary latent features [20].”
• Social GAN: Socially Acceptable Trajectories with Generative Adversarial Networks (2018, methodology): “d [5,6,25,26,30,31,36,46]. Tools popular in economics have also been used such as the Discrete Choice framework by Antonini et. al. [2]. Treuille et. al. [42] use continuum dynamics, and Wang et. al. [44], Tay et. al. [41] use Gaussian processes. Such functions have also been used to study stationar”
• Time series forecasting for nonlinear and non-stationary processes: a review and comparative study (2015, unknown): “…has made forecasting of the evolution of complex biological, physical, and engineering system dynamics (Lang et al., 2007; Duy and Peters, 2008; Wang et al., 2008)—crucial for their preventative control—possible, and 0740-817X C© 2015 “IIE” predictive analytics is emerging as one of the most…”
• DESIRE: Distant Future Prediction in Dynamic Scenes with Interacting Agents (2017, methodology): “Classical methods Path prediction problems have been studied extensively with different approaches such as Kalman filters [18], linear regressions [29] to non-linear Gaussian Process regression models [49, 33, 34, 48], autoregressive models [2] and time-series analysis [32].”
• Neural state machine for character-scene interactions (2019, unknown): “To cope with the problem with template-based approaches, techniques to apply kernel-based approaches are explored for synthesizing novel character movements [Mukai and Kuriyama 2005; Rose et al. 1998; Rose III et al. 2001; Wang et al. 2008].”
• Abstraction Levels for Robotic Imitation: Overview and Computational Approaches (2010, methodology): “Many different techniques can be used by the robot to learn about its own body [41, 84, 108, 129, 144].”
• HuMoR: 3D Human Motion Model for Robust Pose Estimation (2021, methodology): “Early sophisticated motion models for pose tracking used a variety of approaches, including mixtures-of-Gaussians [38], linear embeddings of periodic motion [68, 93, 83], nonlinear embeddings [23], and nonlinear autoregressive models [88, 97, 92, 74].”