2017/09/25 by Mark Pfeiffer, Pfeiffer, Mark, Giuseppe Paolo +9 · 2 citations
Computer Science · Engineering · #Autonomous Vehicle Technology and Safety #FOS: Computer and information sciences #Human Pose and Action Recognition #Robotics (cs.RO) #Video Surveillance and Tracking Methods
paper · pdf · doi:10.48550/arxiv.1709.08528
openalex publication_date 2017/09/25 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
This paper reports on a data-driven, interaction-aware motion prediction\napproach for pedestrians in environments cluttered with static obstacles. When\nnavigating in such workspaces shared with humans, robots need accurate motion\npredictions of the surrounding pedestrians. Human navigation behavior is mostly\ninfluenced by their surrounding pedestrians and by the static obstacles in\ntheir vicinity. In this paper we introduce a new model based on Long-Short Term\nMemory (LSTM) neural networks, which is able to learn human motion behavior\nfrom demonstrated data. To the best of our knowledge, this is the first\napproach using LSTMs, that incorporates both static obstacles and surrounding\npedestrians for trajectory forecasting. As part of the model, we introduce a\nnew way of encoding surrounding pedestrians based on a 1d-grid in polar angle\nspace. We evaluate the benefit of interaction-aware motion prediction and the\nadded value of incorporating static obstacles on both simulation and real-world\ndatasets by comparing with state-of-the-art approaches. The results show, that\nour new approach outperforms the other approaches while being very\ncomputationally efficient and that taking into account static obstacles for\nmotion predictions significantly improves the prediction accuracy, especially\nin cluttered environments.\n