Video anomaly detection using deep incremental slow feature analysis network

Abstract

Existing anomaly detection (AD) approaches rely on various hand‐crafted representations to represent video data and can be costly. The choice or designing of hand‐crafted representation can be difficult when faced with a new dataset without prior knowledge. Motivated by feature learning, e.g. deep leaning and the ability to directly learn useful representations and model high‐level abstraction from raw data, the authors investigate the possibility of using a universal approach. The objective is learning data‐driven high‐level representation for the task of video AD without relying on hand‐crafted representation. A deep incremental slow feature analysis (D‐IncSFA) network is constructed and applied to directly learning progressively abstract and global high‐level representations from raw data sequence. The D‐IncSFA network has the functionalities of both feature extractor and anomaly detector that make AD completion in one step. The proposed approach can precisely detect global anomaly such as crowd panic. To detect local anomaly, a set of anomaly maps, produced from the network at different scales, is used. The proposed approach is universal and convenient, working well in different types of scenarios with little human intervention and low memory and computational requirements. The advantages are validated by conducting extensive experiments on different challenge datasets.

An Overview of Deep Learning Based Methods for Unsupervised and Semi-Supervised Anomaly Detection in Videos

“…The second PCA is applied on the derivative of the normalized input to evaluate the flow features. To achieve a computationally tractable solution, a two-layer localized SFA architecture is proposed by authors [67] for the task of online slow feature extraction and consequent anomaly detection.…”

Distribution Based Crowd Abnormality Detection

“…In the similar work, they shows very interesting description of the spatial saliency on the basis of joint information among the characteristics and the background/foreground classes. Recently, few attribute based models, DL (Deep Learning) methods and the measure based frameworks have been introduced for the identification of abnormal behavior [24]- [29]. In [28,29], the crowd emotions as well as mid-level data is utilized to fill up the space among low level of appearance/motion characteristics and higher level of the crowd behaviors.…”

“…In [25], the measure to catch the crowd motion commotion for the identification of abnormality task is showed. On the other side, the DL methods [24], [26], and [27] normally utilize the learning networks like CNN, PCAnet and IncSFA to remove the semantic data from crowd motion. By integrating the semantic data with the various low-level of visual features are as optical flows and oriented gradients, these techniques recognize the unusual behaviors which is much more accurate.…”

Abnormal event detection in crowded scenes using histogram of oriented contextual gradient descriptor

“…In the first stage, most normal patches were rejected by a small stack of an auto-encode, and a deep convolutional neural network (CNN) was applied to extract the discriminative features for the final decision. Hu et al [30] proposed a deep incremental slow features analysis (D-IncSFA) network to learn the slow features in a scene. Feng et al [31] proposed a deep Gaussian mixture model (D-GMM) network to model normal events.…”