Towards Personalized Pedestrian Route Recommendation Based on Implicit Visual Preference

Abstract

Walking is an everyday, healthy, and eco-friendly mode of trans-portation. The visual environment of roads is crucial for pedestrians' walkingexperiences. However, limited research has explored modeling pedestrians'visual preferences for road environments. In this work-in-progress paper, wepropose a personalized pedestrian route recommender system based on im-plicit visual preferences. Our work primarily consists of three parts: 1. Inves-tigate the relationship between human eye movement and visual preferencestowards the environment. 2. Explore route recommender system methodsbased on pedestrians' implicit visual preferences. 3. Dynamically adjust andupdate recommended results in response to pedestrians' real-time changesin eye movement. This recommender system framework based on implicitvisual preferences also holds significant potential for numerous other do-mains, such as supermarkets, museums, and more.