Atomic Scenes for Scalable Traffic Scene Recognition in Monocular Videos

/in /by

Atomic Scenes for Scalable Traffic Scene Recognition in Monocular VideosPublication Date: March 7, 2016

Event: CVPR 2016

Reference: 10.1109/WACV.2016.7477609

Authors: Chao-Yeh Chen, Wongun Choi, Manmohan Chandraker

Abstract: We propose a novel framework for monocular traffic scene recognition, relying on a decomposition into high-order and atomic scenes to meet those challenges. High-order scenes carry semantic meaning useful for AWS applications, while atomic scenes are easy to learn and represent elemental behaviors based on 3D localization of individual traffic participants. We propose a novel hierarchical model that captures co-occurrence and mutual-exclusion relationships while incorporating both low-level trajectory features and high-level scene features, with parameters learned using a structured support vector machine. We propose efficient inference that exploits the structure of our model to obtain real-time rates.

Publication Link: https://ieeexplore.ieee.org/document/7477609