AutoLTS: Automating Cycling Stress Assessment via Contrastive Learning and Spatial Post-processing
This work addresses a bottleneck in cycling infrastructure planning and route recommendation by automating stress assessment, though it is incremental as it builds on existing deep learning and image analysis methods.
The paper tackled the problem of slow and data-intensive cycling stress assessment by proposing a deep learning framework using street-view images, achieving accurate, fast, and large-scale assessments on a dataset of 39,153 road segments in Toronto.
Cycling stress assessment, which quantifies cyclists' perceived stress imposed by the built environment and motor traffics, increasingly informs cycling infrastructure planning and cycling route recommendation. However, currently calculating cycling stress is slow and data-intensive, which hinders its broader application. In this paper, We propose a deep learning framework to support accurate, fast, and large-scale cycling stress assessments for urban road networks based on street-view images. Our framework features i) a contrastive learning approach that leverages the ordinal relationship among cycling stress labels, and ii) a post-processing technique that enforces spatial smoothness into our predictions. On a dataset of 39,153 road segments collected in Toronto, Canada, our results demonstrate the effectiveness of our deep learning framework and the value of using image data for cycling stress assessment in the absence of high-quality road geometry and motor traffic data.