Intention Recognition in Real-Time Interactive Navigation Maps

TL;DR

IntentRec4Maps tackles real-time intention recognition for interactive navigation maps by combining a Google Maps Platform based map facade with a MirroringACS_online online recognizer. The approach computes ideal routes for each possible intent and comparison with observation-driven routes to derive a likelihood $P(loc|Obs)$ via a score $ε ∈ [0,1]$. Experiments on a 100-problem dataset show baseline Google Maps route data yields the highest accuracy, with MapBox moderately lower and an LLM-based route planner performing worst, highlighting the practicality and current limits of LLMs for route extraction. The work demonstrates a novel integration of real-time intention recognition into interactive maps and outlines future directions for robustness to noise, irrational or adversarial observations, and long-distance tasks.

Abstract

In this demonstration, we develop IntentRec4Maps, a system to recognise users' intentions in interactive maps for real-world navigation. IntentRec4Maps uses the Google Maps Platform as the real-world interactive map, and a very effective approach for recognising users' intentions in real-time. We showcase the recognition process of IntentRec4Maps using two different Path-Planners and a Large Language Model (LLM). GitHub: https://github.com/PeijieZ/IntentRec4Maps

Figures (3)

• Figure 1: IntentRec4Maps Overview.
• Figure 2: IntentRec4Maps Recognition Process.
• Figure 3: Haversine example.