Sökning: onr:"swepub:oai:DiVA.org:mdh-66496" >
Modeling interpreta...
Modeling interpretable social interactions for pedestrian trajectory
-
- Liu, Q. (författare)
- Department of Civil Engineering, McGill University, Montreal, Quebec, H3A 0C3, Canada. Department of Traffic Engineering, Shanghai Jiao Tong University, Shanghai, 200240, China
-
- Shi, Xiaodan (författare)
- Mälardalens universitet,Framtidens energi,Center for Spatial Information Science, The University of Tokyo, Kashiwa, Japan
-
- Jiang, R. (författare)
- Center for Spatial Information Science, The University of Tokyo, Kashiwa, Japan. Information Technology Center, The University of Tokyo, Kashiwa, Japan
-
visa fler...
-
- Zhang, H. (författare)
- School of Urban Planning and Design, Peking University, Shenzhen, Guangdong, 518055, China
-
- Lu, L. (författare)
- Department of Traffic Engineering, Shanghai Jiao Tong University, Shanghai, 200240, China
-
- Shibasaki, R. (författare)
- Center for Spatial Information Science, The University of Tokyo, Kashiwa, Japan
-
visa färre...
-
Department of Civil Engineering, McGill University, Montreal, Quebec, H3A 0C3, Canada Department of Traffic Engineering, Shanghai Jiao Tong University, Shanghai, 200240, China Framtidens energi (creator_code:org_t)
- Elsevier Ltd, 2024
- 2024
- Engelska.
-
Ingår i: Transportation Research Part C. - : Elsevier Ltd. - 0968-090X .- 1879-2359. ; 162
- Relaterad länk:
-
https://doi.org/10.1...
-
visa fler...
-
https://urn.kb.se/re...
-
https://doi.org/10.1...
-
visa färre...
Abstract
Ämnesord
Stäng
- The abilities to understand pedestrian social interaction behaviors and to predict their future trajectories are critical for road safety, traffic management and more broadly autonomous vehicles and robots. Social interactions are intuitively heterogeneous and dynamic over time and circumstances, making them hard to explain. In this paper, we creatively investigate modeling interpretable social interactions for pedestrian trajectory, which is not considered by the existing trajectory prediction research. Moreover, we propose a two-stage methodology for interaction modeling - “mode extraction” and “mode aggregation”, and develop a long short-term memory (LSTM)-based model for long-term trajectory prediction, which naturally takes into account multi-types of social interactions. Different from previous models that do not explain how pedestrians interact socially, we extract latent modes that represent social interaction types which scales to an arbitrary number of neighbors. Extensive experiments over two public datasets have been conducted. The quantitative and qualitative results demonstrate that our method is able to capture the multi-modality of human motion and achieve better performance under specific conditions. Its performance is also verified by the interpretation of predicted modes, of which the results are in accordance with common sense. Besides, we have performed sensitivity analysis on the crucial hyperparameters in our model. Code is available at: https://github.com/xiaoluban/Modeling-Interpretable-Social-Interactions-for-Pedestrian-Trajectory.
Ämnesord
- NATURVETENSKAP -- Data- och informationsvetenskap (hsv//swe)
- NATURAL SCIENCES -- Computer and Information Sciences (hsv//eng)
Nyckelord
- Deep learning
- Explainability and comprehensibility of AI
- Interpretable social interactions
- Long short-term memory (LSTM)
- Multi-modality
- Trajectory prediction
- Brain
- Forecasting
- Motor transportation
- Pedestrian safety
- Sensitivity analysis
- Trajectories
- Interaction behavior
- Interpretable social interaction
- Long short-term memory
- Pedestrian trajectories
- Performance
- Social interactions
Publikations- och innehållstyp
- ref (ämneskategori)
- art (ämneskategori)
Hitta via bibliotek
Till lärosätets databas