SILA: An Incremental Learning Approach for Pedestrian Trajectory Prediction

Abstract

The prediction of pedestrian motion is challenging, especially in crowded roads and intersections. Most of the current approaches apply offline methods to learn motion behaviors, but as a result, they are not able to learn continuously and typically do not generalize well to new environments. This paper presents Similarity-based Incremental Learning Algorithm (SILA) for pedestrian motion prediction with the ability of improving the learned model over the time as data is obtained incrementally. To keep the model size efficient, the motion primitives learned from the new data are compared with the previously known ones, and similar motion primitives are fused while novel motion primitives are added to the model. Results show that the SILA model growth rate is about 1/3 that of an incremental approach that does not fuse motion primitives. SILA is evaluated on different datasets and scenarios including intersections and busy streets. The results show that, even though SILA learns incrementally, it performs comparably to (and sometimes outperforms) state-of-the-art algorithms in pedestrian prediction. Additionally, SILA learning time only depends on the size of the data added incrementally, which makes SILA more efficient in terms of time and space compared to batch learning.

Survey of maps of dynamics for mobile robots

“…The idea of utilization of directional maps for motion prediction has received attention only in recent years. In Habibi et al (2018), the authors propose a context-based approach for pedestrian motion prediction in crowded, urban intersections. They incorporated semantic features from the environment (relative distance to curbside and status of pedestrian traffic lights) in the GP formulation for more accurate predictions of pedestrian trajectories over the same timescale.…”

Incoherent and Online Dictionary Learning Algorithm for Motion Prediction

“…TASNSC was further extended as an incremental learning framework, SILA [20], where the model was incrementally updated when new trajectory samples/datasets were available. SILA used the TASNSC method to separately learn dictionaries/motion primitives on different datasets.…”

“…The recently proposed incremental learning algorithms SILA [20] and SimFUSE [21] have been shown to generate high quality predictions when compared with other state-ofthe-art algorithms. These methods learn the baseline motion primitives using the dictionary learning method ASNSC [9] and perform fusion of the learned motion primitives for performance improvement.…”

Context-aware pedestrian motion prediction

“…Graph Neural Network (GNN) [76] Interaction Graph approaches [6,43,45,48,56,84,88,89] and Graph Attention Networks (GATs) [87] approaches [54,55,59].…”

“…Therefore, the knowledge that social relational model learns from urban crowd situation as in a top-down view can be transferred to the vehicle-pedestrian situation, making partial use of the former situation as a prior knowledge. A recent trend tackles incremental adaptive learning [89], which applies knowledge transfer to novel scenes without fully re-training on new data. Given that this study aimed at reducing graph design complexity and introduced the online self-growth, additional contributions would be added to the existing paradigm.…”