An Anomaly Comprehension Neural Network for Surveillance Videos on Terminal Devices

Yuan Cheng, Guangtai Huang, Peining Zhen, Bin Liu, Hai-Bao Chen, Ngai Wong, Hao Yu

Abstract

Anomaly comprehension in surveillance videos is more challenging than detection. This work introduces the design of a lightweight and fast anomaly comprehension neural network. For comprehension, a spatio-temporal LSTM model is developed based on the structured, tensorized time-series features extracted from surveillance videos. Deep compression of network size is achieved by tensorization and quantization for the implementation on terminal devices. Experiments on large-scale video anomaly dataset UCF-Crime demonstrate that the proposed network can achieve an impressive inference speed of 266 FPS on a GTX-1080Ti GPU, which is 4.29 faster than ConvLSTM-based method; a 3.34% AUC improvement with 5.55% accuracy niche versus the 3D-CNN based approach; and at least 15k× parameter reduction and 228× storage compression over the RNN-based approaches. Moreover, the proposed framework has been realized on an ARM-core based IOT board with only 2.4W power consumption.
BibTeX
@inproceedings{cheng2020anomaly,
  author = {Yuan Cheng and Guangtai Huang and Peining Zhen and Bin Liu and Hai-Bao Chen and Ngai Wong and Hao Yu},
  title = {{An Anomaly Comprehension Neural Network for Surveillance Videos on Terminal Devices}},
  booktitle = {DATE},
  year = {2020},
  doi = {10.23919/date48585.2020.9116533},
}

Loading related content...

Figure