Graph Convolutional Label Noise Cleaner: Train a Plug-And-Play Action Classifier for Anomaly Detection

By Jia-Xing Zhong, Nannan Li, Weijie Kong, Shan Liu, Thomas H. Li, Ge Li. Computer Vision and Pattern Recognition, 2019

535 citations recorded by the source corpus.

Video anomaly detection under weak labels is formulated as a typical multiple-instance learning problem in previous works. In this paper, we provide a new perspective, i.e., a supervised learning task under noisy labels. In such a viewpoint, as long as cleaning away label noise, we can directly apply fully supervised action classifiers to weakly supervised anomaly detection, and take maximum advantage of these well-developed classifiers. For this purpose, we devise a graph convolutional network to correct noisy labels. Based upon feature similarity and temporal consistency, our network propagates supervisory signals from high-confidence snippets to low-confidence ones. In this manner, the network is capable of providing cleaned supervision for action classifiers. During the test phase, we only need to obtain snippet-wise predictions from the action classifier without any extra post-processing. Extensive experiments on 3 datasets at different scales with 2 types of action classifiers demonstrate the efficacy of our method. Remarkably, we obtain the frame-level AUC score of 82.12% on UCF-Crime.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/a03bda078490e8ee991a1f86b53f27df7cf93a14

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

In Computer Vision and Pattern Recognition

Citations in the corpus

MIST: Multiple Instance Self-Training Framework for Video Anomaly Detection (2021, methodology): “The encoder-based methods [38, 35] train both the feature encoder and classiﬁer simultaneously.”
Image Classification with Deep Learning in the Presence of Noisy Labels: A Survey (2019, background): “[150] formulates video anomaly detection as a classification with label noise”
DualGraph: A graph-based method for reasoning about label noise (2021, background): “Recent approaches [10, 39, 40] are proposed to exploit graph neural network in the ﬁeld of label noise task.”
Self-Training Multi-Sequence Learning with Transformer for Weakly Supervised Video Anomaly Detection (2022, methodology): “We report the results on ShanghaiTech (Zhong et al. 2019) in Table 1.”
Generalized Out-of-Distribution Detection: A Survey (2021, unknown): “Techniques such as sample re-weighting [265] and label cleaning methods [266], [267] have also shown their effectiveness for the task.”
Anomaly Detection in Video via Self-Supervised and Multi-Task Learning (2020, unknown): “[48] and continued by others [65] less related to our study.”
Weakly-supervised Video Anomaly Detection with Robust Temporal Feature Magnitude Learning (2021, methodology): “Moreover, our method unifies the representation learning and anomaly score learning by an `2-norm-based temporal feature ranking loss, enabling better separation between normal and abnormal feature representations, improving the exploration of weak labels compared to previous MIL methods [56, 62, 66”
Not only Look, but also Listen: Learning Multimodal Violence Detection under Weak Supervision (2020, methodology): “Inspired by the GCN for video understanding [51,41,50], we define the holistic relation matrix by feature similarity prior,”
Global-Local GCN: Large-Scale Label Noise Cleansing for Face Recognition (2020, unknown): “[57] deploy GCN for anomaly detection.”