Learning Regularity in Skeleton Trajectories for Anomaly Detection in Videos

By Romero Morais, Vuong Le, Truyen Tran, Budhaditya Saha, Moussa Reda Mansour, Svetha Venkatesh. Computer Vision and Pattern Recognition, 2019

321 citations recorded by the source corpus.

Appearance features have been widely used in video anomaly detection even though they contain complex entangled factors. We propose a new method to model the normal patterns of human movements in surveillance video for anomaly detection using dynamic skeleton features. We decompose the skeletal movements into two sub-components: global body movement and local body posture. We model the dynamics and interaction of the coupled features in our novel Message-Passing Encoder-Decoder Recurrent Network. We observed that the decoupled features collaboratively interact in our spatio-temporal model to accurately identify human-related irregular events from surveillance video sequences. Compared to traditional appearance-based models, our method achieves superior outlier detection performance. Our model also offers “open-box” examination and decision explanation made possible by the semantically understandable features and a network architecture supporting interpretability.

Subfields: Artificial Intelligence, Biomedical Engineering, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/db20d81d40243d66ff90f11b5c6f058d43d3701f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Truyen Tran
2. Svetha Venkatesh
3. Antonio Jimeno-Yepes
4. Vuong Le
5. Qing Zhang
6. Guiqing Li
7. Xu Zhong
8. Zhian Liu

In Computer Vision and Pattern Recognition

1. Antonio Jimeno-Yepes
2. Svetha Venkatesh
3. Vuong Le
4. Qing Zhang
5. Guiqing Li
6. Xu Zhong
7. Zhian Liu
8. Yongwei Nie

Citations in the corpus

• A Hybrid Video Anomaly Detection Framework via Memory-Augmented Flow Reconstruction and Flow-Guided Frame Prediction (2021, background): “Recently, several works [50, 46, 35] are proposed to combine these two paradigms in a hybrid manner.”
• Neural Reasoning, Fast and Slow, for Video Question Answering (2019, background): “e use of the standard bi-LSTM with GloVe word embedding. The visual representation process has been studied intensively in recent years, in particular for action recognition [38], trajectory modeling [25] and video prediction [24]. Video QA represents a challenging class compared to action recogniti”
• Image-based table recognition: data, model, and evaluation (2019, methodology): “Our model differs from other existing EDD architectures [35], [36], where the dual decoders are independent from each other.”