Pedestrian Trajectory Prediction Using Dynamics-based Deep Learning

Pedestrian trajectory prediction plays an important role in autonomous driving systems and robotics. Recent work utilizing prominent deep learning models for pedestrian motion prediction makes limited a priori assumptions about human movements, resulting in a lack of explainability and explicit constraints enforced on predicted trajectories. We present a dynamics-based deep learning framework with a novel asymptotically stable dynamical system integrated into a Transformer-based model. We use an asymptotically stable dynamical system to model human goal-targeted motion by enforcing the human walking trajectory, which converges to a predicted goal position, and to provide the Transformer model with prior knowledge and explainability. Our framework features the Transformer model that works with a goal estimator and dynamical system to learn features from pedestrian motion history. The results show that our framework outperforms prominent models using five benchmark human motion datasets.

Authors

• Honghui Wang
• Weiming Zhi
• Gustavo Batista
• Rohitash Chandra

Keywords

• Deep learning
• Legged locomotion
• Pedestrians
• Predictive models
• Benchmark testing
• Transformers
• Trajectory
• Trajectory Prediction
• Pedestrian Trajectory
• Pedestrian Trajectory Prediction
• System Dynamics
• Stability Of System
• Deep Learning Models
• Asymptotically Stable
• Human Movement
• Transformer Model
• Human Motion
• Human Walking
• Explicit Constraints
• Human Trajectory
• Neural Network
• Collision
• Convolutional Network
• Straight Line
• Unsupervised Learning
• Long Short-term Memory
• Graph Convolutional Network
• Stable Attractor
• Lower Triangular
• Positive Definite Matrix
• Recurrent Neural Network
• Temporal Sequence
• Equilibrium Point
• End Position
• Social Forces
• Similar Trajectories