Pedestrian Path Prediction Using Markov Decision Processes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Predicting pedestrian paths is challenging due to the difficulty in determining future movements, especially when pedestrians stop or change direction, as current trajectory-based predictions are insufficient and fail to account for environmental features and traffic rules.
Innovation Solution
A system comprising a sensor component, modeling component, prediction component, and interface component that identifies pedestrians and environment features, generates models using Markov Decision Processes (MDPs) to infer pedestrian goals and navigation preferences, and predicts paths with associated risk scores, accounting for traffic rules and environmental factors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If trajectory-based prediction is used, then the prediction method is simple, but the prediction accuracy is insufficient
Solution Approach 1:
The patent transforms the prediction approach from simple trajectory extrapolation to a comprehensive model that incorporates multiple parameters including environmental features, traffic rules, pedestrian goals, and navigation preferences. This parameter enrichment resolves the contradiction by making the prediction system more accurate while managing the increased complexity through structured modeling.
Solution Approach 2:
The patent introduces an intermediary modeling component that acts as a bridge between raw sensor data and prediction outputs. This mediator processes environmental features, infers pedestrian goals, and generates predicted paths, thereby improving prediction accuracy without directly exposing the complexity of the underlying algorithms to the user.
2Measurement precision
If environmental features and traffic rules are incorporated, then prediction accuracy improves, but system complexity increases
Solution Approach 1:
The patent segments the complex prediction system into distinct functional modules: sensor component for data collection, modeling component for goal inference and path generation, and prediction component for output generation. This segmentation manages system complexity by organizing multiple environmental and traffic rule factors into manageable, modular components that can be processed independently.
3Measurement precision
If Markov Decision Process modeling is used, then pedestrian goal inference accuracy improves, but computational requirements increase
Solution Approach 1:
The patent applies partial MDP modeling by focusing computational resources on inferring pedestrian goals and immediate navigation preferences rather than modeling all possible future states. This partial action approach maintains high goal inference accuracy while reducing computational energy requirements by avoiding exhaustive state-space exploration.
Data Source
AI summary
Systems and techniques for pedestrian path predictions are disclosed herein. For example, an environment, features of the environment, and pedestrians within the environment may be identified. Models for the pedestrians may be generated based on features of the environment. A model may be indicative of goals of a corresponding pedestrian and predicted paths for the corresponding pedestrian. Pedestrian path predictions for the pedestrians may be determined based on corresponding predicted paths. A pedestrian path prediction may be indicative of a probability that the corresponding pedestrian will travel a corresponding predicted path. Pedestrian path predictions may be rendered for the predicted paths, such as using different colors or different display aspects, thereby enabling a driver of a vehicle to be presented with information indicative of where a pedestrian is likely to travel.