@article{Wang2025, 
author = {Zihe Wang and Haiyang Yu and Changxin Chen and Zhiyong Cui and Yufeng Bi and Yilong Ren and Zijian Wang and Delan Kong and Jing Tian and Shoutong Yuan and Zhiqiang Li},
title = {MoTIF: An end-to-end multimodal road traffic scene understanding foundation model},
year = {2025},
journal = {Communications in Transportation Research},
volume = {5},
number = {4},
pages = {100227},
keywords = {Road traffic, Scene understanding, Multimodal foundation model, Fine-tuning},
url = {https://www.sciopen.com/article/10.1016/j.commtr.2025.100227},
doi = {10.1016/j.commtr.2025.100227},
abstract = {Video-based road intelligent detection constitutes a critical component in modern intelligent transportation systems, serving as a crucial role for comprehensive transportation planning and emergency traffic management. Current traffic scene perception methodologies relying on conventional deep learning architectures present inherent limitations, including heavy dependence on extensive manual annotations of specific traffic scenarios and predefined rule configurations. These approaches demonstrate constrained semantic representation capacity and limited generalizability across heterogeneous traffic scenarios. To address these challenges, this study proposes a novel end-to-end multimodal foundation model architecture that jointly generates dynamic traffic event detection outcomes and semantic-rich contextual descriptions. Through integration of low-rank adaptation (LoRA) and prompt fine-tuning as parameter-efficient fine-tuning strategies, we develop the multimodal road traffic scene understanding foundation model (MoTIF), which establishes cross-modal alignment between visual patterns and textual semantics. This framework demonstrates enhanced capability in extracting salient traffic targets and generating hierarchical scene representations, significantly improving automated detection efficiency in road video analytics. Notably, MoTIF exhibits contextual reasoning capabilities for implicit traffic event interpretation. Extensive evaluations on two real-world datasets encompassing urban road intersection scenarios in Tianjin and highway monitoring systems in Shandong Province reveal that MoTIF achieves superior performance metrics: 65.81 average score on multimodal scene understanding assessment and 83.33% event detection accuracy, outperforming mainstream benchmarks in both precision and computational efficiency. This research advances multimodal learning paradigms for intelligent transportation systems while providing practical insights for adaptive traffic management applications.}
}