Computer ScienceCommunications in Transportation Research
Vision Language Model–Guided Semantic Distillation Network for Pedestrian Crossing Intention Prediction
Y. Li, C. Wang, et al.
Can autonomous vehicles anticipate when a pedestrian is about to cross without juggling countless data streams? Research conducted by Yuanzhe Li, Cheng Wang, Zhuoqi Zeng, and Steffen Müller introduces VLM-GSDN, a semantic distillation network that learns from vision-language descriptions during training but needs only RGB images and pedestrian motion for inference. Tested on JAAD and PIE, the approach achieves competitive state-of-the-art performance, promising cross-dataset generalization, and robustness across longer prediction horizons.
Related Publications
Explore these studies to deepen your understanding
Adjacent work that informs or extends this paper's methodology and findings.
Computer Science
YesBut: A High-Quality Annotated Multimodal Dataset for evaluating Satire Comprehension capability of Vision-Language Models
A. Nandy, Y. Agarwal, et al.
Business
Abstract or concrete? The effects of language style and service context on continuous usage intention for AI voice assistants
H. Lan, X. Tang, et al.
Medicine and Health
Vision-language foundation model for echocardiogram interpretation
M. Christensen, M. Vukadinovic, et al.
Linguistics and Languages
Specific disruption of the ventral anterior temporo-frontal network reveals key implications for language comprehension and cognition
E. Kourtidou, D. Kasselimis, et al.
