Towards Accurate Vehicle Behaviour Classification With Multi-Relational Graph Convolutional Networks
This work addresses the problem of accurate on-road vehicle behavior understanding for applications like autonomous driving and traffic monitoring, representing an incremental improvement with a novel method for a known bottleneck.
The paper tackles vehicle behavior classification from monocular video by proposing a pipeline that uses spatial information encoded by a Multi-Relational Graph Convolutional Network and temporal processing with a recurrent network, achieving high fidelity on diverse datasets including European, Chinese, and Indian scenes and enabling seamless model transfer without re-annotation or retraining.
Understanding on-road vehicle behaviour from a temporal sequence of sensor data is gaining in popularity. In this paper, we propose a pipeline for understanding vehicle behaviour from a monocular image sequence or video. A monocular sequence along with scene semantics, optical flow and object labels are used to get spatial information about the object (vehicle) of interest and other objects (semantically contiguous set of locations) in the scene. This spatial information is encoded by a Multi-Relational Graph Convolutional Network (MR-GCN), and a temporal sequence of such encodings is fed to a recurrent network to label vehicle behaviours. The proposed framework can classify a variety of vehicle behaviours to high fidelity on datasets that are diverse and include European, Chinese and Indian on-road scenes. The framework also provides for seamless transfer of models across datasets without entailing re-annotation, retraining and even fine-tuning. We show comparative performance gain over baseline Spatio-temporal classifiers and detail a variety of ablations to showcase the efficacy of the framework.