Fine-Tuning a Pre-Trained CAE for Deep One Class Anomaly Detection

EasyChair Preprint 4277

Abstract

In recent years, abnormal event detection in video surveillance is become avery important task mainly treated by deep learning methods taken in account many challenges. However, these methods still not trained on an anomaly detection based objective which proves their ineffectiveness in such problem . In this paper, we propose an unsupervised method based on new architecture for deep one class of convolutional auto-encoders (CAEs) for representing a compact spatio-temporal feature for anomaly detection. Our CAEs are constructed by added deconvolutions layers to the convolutions layers of VGG 16. Then, we train our CAEs for a one-class training objective by finetuning our model to properly exploit the richness of the dataset with which CNN was trained. The first CAE is trained on the original frames to extract a good descriptor of shapes and the second CAE is learned using optical flow representations to provide a strength description of motion between frames. For this purpose, we define two loss functions , compactness loss and representativeness loss for training our CAEs architectures not only to maximize the inter-classes distance and to minimize the intra-class distance but also to ensure the tightness and the representativeness of features of normal images. We reduce features dimensions by applying a PCA ( Principal Component Analyser ) in order to combine our two descriptors with a Gaussian classifier for abnormal spatio-temporal events detection. Our method has a high performance in terms of reliability and accuracy. It achieved abnormal event detection with good efficiency in challenging datasets compared to state-of-the-art methods

Keyphrases: Deep one class, anomaly detection, convolutional auto-encoder, deep learning, fine-tuning

@booklet{EasyChair:4277,
  author    = {Slim Hamdi and Hichem Snoussi and Mohamed Abid},
  title     = {Fine-Tuning a Pre-Trained CAE for Deep One Class Anomaly Detection},
  howpublished = {EasyChair Preprint 4277},
  year      = {EasyChair, 2020}}