LCrowdV: Generating Labeled Videos for Simulation-based Crowd Behavior Learning
This addresses the data scarcity problem for researchers developing crowded scene understanding algorithms, though it is an incremental improvement over existing simulation-based data generation methods.
The authors tackled the problem of limited labeled crowd video data by developing LCrowdV, a procedural framework that generates arbitrary numbers of automatically labeled crowd videos with various parameters like density, behavior, and lighting. They demonstrated that using LCrowdV improves pedestrian detection and crowd behavior classification accuracy compared to prior datasets.
We present a novel procedural framework to generate an arbitrary number of labeled crowd videos (LCrowdV). The resulting crowd video datasets are used to design accurate algorithms or training models for crowded scene understanding. Our overall approach is composed of two components: a procedural simulation framework for generating crowd movements and behaviors, and a procedural rendering framework to generate different videos or images. Each video or image is automatically labeled based on the environment, number of pedestrians, density, behavior, flow, lighting conditions, viewpoint, noise, etc. Furthermore, we can increase the realism by combining synthetically-generated behaviors with real-world background videos. We demonstrate the benefits of LCrowdV over prior lableled crowd datasets by improving the accuracy of pedestrian detection and crowd behavior classification algorithms. LCrowdV would be released on the WWW.