Real-World Anomaly Detection in Surveillance Videos

Обнаружение аномалий в видеозаписях с камер наблюдения в реальных условиях
Waqas Sultani, Chen Chen, Mubarak A. Shah
2018-06-01

multiple instance learningreal-world anomaly detectionsurveillance videostemporal smoothness constraintsweakly labeled videos
Surveillance videos are able to capture a variety of realistic anomalies. In this paper, we propose to learn anomalies by exploiting both normal and anomalous videos. To avoid annotating the anomalous segments or clips in training videos, which is very time consuming, we propose to learn anomaly through the deep multiple instance ranking framework by leveraging weakly labeled training videos, i.e. the training labels (anomalous or normal) are at video-level instead of clip-level. In our approach, we consider normal and anomalous videos as bags and video segments as instances in multiple instance learning (MIL), and automatically learn a deep anomaly ranking model that predicts high anomaly scores for anomalous video segments. Furthermore, we introduce sparsity and temporal smoothness constraints in the ranking loss function to better localize anomaly during training. We also introduce a new large-scale first of its kind dataset of 128 hours of videos. It consists of 1900 long and untrimmed real-world surveillance videos, with 13 realistic anomalies such as fighting, road accident, burglary, robbery, etc. as well as normal activities. This dataset can be used for two tasks. First, general anomaly detection considering all anomalies in one group and all normal activities in another group. Second, for recognizing each of 13 anomalous activities. Our experimental results show that our MIL method for anomaly detection achieves significant improvement on anomaly detection performance as compared to the state-of-the-art approaches. We provide the results of several recent deep learning baselines on anomalous activity recognition. The low recognition performance of these baselines reveals that our dataset is very challenging and opens more opportunities for future work. The dataset is available at: http://crcv.ucf.edu/projects/real-world.
1
A deep multiple-instance ranking framework learns anomaly scores from weakly labeled videos, avoiding clip-level annotation of anomalous segments.
2
Normal and anomalous videos are modeled as bags and video segments as instances, with sparsity and temporal smoothness constraints improving anomaly localization.
3
The authors introduce a 128-hour dataset containing 1,900 long, untrimmed surveillance videos with 13 realistic anomaly types and normal activities.
4
The dataset supports both general anomaly detection and recognition of individual anomalous activities.
5
The proposed MIL method significantly outperforms state-of-the-art approaches for anomaly detection, while baseline recognition models perform poorly, demonstrating the dataset’s difficulty.

real-world surveillance videos containing normal activities and 13 types of realistic anomalies

weakly supervised temporal anomaly detection and localization, including anomaly ranking across video segments and recognition of anomalous activities

Publication Details
Publication Date
2018-06-01
Journal
IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings
Publisher
Institute of Electrical and Electronics Engineers
ISSN
1063-6919
Cited by
3006
Access Type
Author Information
Authors
Waqas Sultani
Chen Chen
Mubarak A. Shah
Explore further
Open the scid.ai AI chat with a ready-made request: it will find papers on a similar topic and help build a literature review.
Find similar papers in the chat →
Make a presentation
100%