Human in Events: A Large-Scale Benchmark for Human-centric Video Analysis in Complex Events

TL;DR

HiEve introduces a large-scale, crowd-centric video dataset with synchronized annotations for tracks, poses, and actions to benchmark human-centric analysis in complex events. It demonstrates that HiEve is challenging across tracking, pose estimation, and action recognition, and validates this with extensive baselines, including two cross-annotation approaches that exploit pose-action information during training. The paper also provides an online evaluation server, detailed statistics, and novel weighted metrics to better assess performance in crowded scenes, along with ablation analyses and transfer observability. Collectively, HiEve enables more realistic evaluation and fosters development of robust methods for complex human-centric video analytics.

Abstract

Along with the development of modern smart cities, human-centric video analysis has been encountering the challenge of analyzing diverse and complex events in real scenes. A complex event relates to dense crowds, anomalous individuals, or collective behaviors. However, limited by the scale and coverage of existing video datasets, few human analysis approaches have reported their performances on such complex events. To this end, we present a new large-scale dataset with comprehensive annotations, named Human-in-Events or HiEve (Human-centric video analysis in complex Events), for the understanding of human motions, poses, and actions in a variety of realistic events, especially in crowd & complex events. It contains a record number of poses (>1M), the largest number of action instances (>56k) under complex events, as well as one of the largest numbers of trajectories lasting for longer time (with an average trajectory length of >480 frames). Based on its diverse annotation, we present two simple baselines for action recognition and pose estimation, respectively. They leverage cross-label information during training to enhance the feature learning in corresponding visual tasks. Experiments show that they could boost the performance of existing action recognition and pose estimation pipelines. More importantly, they prove the widely ranged annotations in HiEve can improve various video tasks. Furthermore, we conduct extensive experiments to benchmark recent video analysis approaches together with our baseline methods, demonstrating HiEve is a challenging dataset for human-centric video analysis. We expect that the dataset will advance the development of cutting-edge techniques in human-centric analysis and the understanding of complex events. The dataset is available at http://humaninevents.org

Figures (46)

• Figure 1: An example of the collection workflow of our HiEve dataset under street scene, where each scene contains videos captured at different sites with different types of events happening.
• Figure 2: Samples of different actions from our training set and testing set.
• Figure 3:
• Figure 4:
• Figure 6: The classification of events. They are divided into three event categories.