The Role of Data-Driven Priors in Multi-Agent Crowd Trajectory Estimation

Authors: Gang Qiao, Sejong Yoon, Mubbasir Kapadia, Vladimir Pavlovic

AAAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Various combinations of priors and optimization algorithms are evaluated in comprehensive simulated experiments. Our experimental results reveal important insights, including the significance of the global flow prior and the lesser-than-expected influence of datadriven collision priors.
Researcher Affiliation Academia 1Rutgers University, 2The College of New Jersey {gq19, mk1353, vladimir}@cs.rutgers.edu, yoons@tcnj.edu
Pseudocode Yes Algorithm 1: Proposed Optimization Framework
Open Source Code No The paper does not provide any explicit statement about releasing source code for the methodology, nor does it include a link to a code repository.
Open Datasets Yes The ground truth trajectories are obtained by running Steer Suite (Singh et al. 2009) library with social force AI (Helbing and Moln ar 1995), and are split into a training and testing sets.
Dataset Splits No The paper mentions splitting data into "training and testing sets" but does not specify a separate validation set or details about its split (e.g., percentages or methods like cross-validation).
Hardware Specification No The paper discusses computational time but does not provide any specific hardware details such as CPU, GPU models, or memory used for running the experiments.
Software Dependencies No The paper mentions tools and models like "Steer Suite (Singh et al. 2009)" and "social force AI (Helbing and Moln ar 1995)" as data sources, and refers to algorithms like "message-passing ADMM (MPA)", "interior point method (IPM)", and "unscented Kalman smoother (UKS)" for optimization. However, it does not specify any software names with version numbers that would be necessary for reproduction.
Experiment Setup Yes We set the parameter as: ut = 1 if the point is actually observed, otherwise ut = 0; Ckn = 1 by assuming homogeneous crowd, Cmv = 2.6m/s, Δt = 1.5s and λ = 1/(σ2 NNΔt2) 108.0.