Trajectory saliency detection using consistency-oriented latent codes from a recurrent auto-encoder
This work tackles the problem of detecting abnormal motion trajectories for applications like alarm triggering or event detection, which is relevant for surveillance and safety systems.
The paper addresses the detection of progressive dynamic saliency from video sequences, specifically focusing on motion-related saliency that evolves over time. It proposes a method that uses a recurrent auto-encoder to learn compact and discriminative latent codes for trajectories, enforcing consistency for normal trajectories. The method detects salient trajectories by measuring the distance of a trajectory's code to a prototype code representing normality, outperforming existing methods on the Alahi 2014 pedestrian trajectory dataset.
In this paper, we are concerned with the detection of progressive dynamic saliency from video sequences. More precisely, we are interested in saliency related to motion and likely to appear progressively over time. It can be relevant to trigger alarms, to dedicate additional processing or to detect specific events. Trajectories represent the best way to support progressive dynamic saliency detection. Accordingly, we will talk about trajectory saliency. A trajectory will be qualified as salient if it deviates from normal trajectories that share a common motion pattern related to a given context. First, we need a compact while discriminative representation of trajectories. We adopt a (nearly) unsupervised learning-based approach. The latent code estimated by a recurrent auto-encoder provides the desired representation. In addition, we enforce consistency for normal (similar) trajectories through the auto-encoder loss function. The distance of the trajectory code to a prototype code accounting for normality is the means to detect salient trajectories. We validate our trajectory saliency detection method on synthetic and real trajectory datasets, and highlight the contributions of its different components. We show that our method outperforms existing methods on several scenarios drawn from the publicly available dataset of pedestrian trajectories acquired in a railway station (Alahi 2014).