WildGraph: Realistic Long-Horizon Trajectory Generation with Limited Sample Size

2024-04-11

SIGSPATIAL/GIS

AI Summary

To address the scarcity of long-term wildlife migration trajectories—only ~60 real-world samples available—this paper proposes a hierarchical generative framework that circumvents privacy, ethical, and technical constraints of field data collection. Methodologically, it introduces the first integration of H3 geospatial indexing with a Recurrent Variational Autoencoder (RVAE): global movement patterns are modeled at coarse spatial-temporal scales, followed by recursive refinement of local path segments. A prototypical network enhances few-shot generalization, while region-occupancy probability priors enable controllable, high-fidelity trajectory synthesis. Evaluated on two real-world migration datasets, our approach significantly improves long-horizon trajectory generation quality, outperforming or matching state-of-the-art methods on FID, Dynamic Time Warping (DTW), and Path Consistency metrics. The framework establishes a scalable, interpretable paradigm for modeling sparse spatiotemporal migration data.

Technology Category

• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Machine Learning: Deep Generative Models & Autoencoders

Application Category

• Web Mining and Content Analysis: Web data generation and simulation
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems

Abstract

Trajectory generation is an important task in movement studies; it circumvents the privacy, ethical, and technical challenges of collecting real trajectories from the target population. In particular, real trajectories in the wildlife domain are scarce as a result of ethical and environmental constraints of the collection process. In this paper, we consider the problem of generating long-horizon trajectories, akin to wildlife migration, based on a small set of real samples. We propose a hierarchical approach to learn the global movement characteristics of the real dataset and recursively refine localized regions. Our solution, WildGraph, discretizes the geographic path into a prototype network of H3 (https://www.uber.com/blog/h3/) regions and leverages a recurrent variational auto-encoder to probabilistically generate paths over the regions, based on occupancy. WildGraph successfully generates realistic months-long trajectories using a sample size as small as 60. Experiments performed on two wildlife migration datasets demonstrate that our proposed method improves the generalization of the generated trajectories in comparison to existing work while achieving superior or comparable performance in several benchmark metrics. Our code is published on the following repository: https://github.com/aliwister/wildgraph.

Problem

• Generate wildlife migration trajectories
• Address ethical data collection constraints
• Improve generalization with small sample sizes

Innovation

• Hierarchical approach for trajectory generation
• Recurrent variational auto-encoder for path generation
• H3 regions for geographic discretization