Future Video Prediction from a Single Frame for Video Anomaly Detection

被引:2
|
作者
Baradaran, Mohammad [1 ]
Bergevin, Robert [1 ]
机构
[1] Univ Laval, 065 Ave Med, Quebec City, PQ G1V 0A6, Canada
关键词
Video anomaly detection; long-term motion modeling; future video prediction; semi-supervised learning; NETWORK;
D O I
10.1007/978-3-031-47969-4_37
中图分类号
TP301 [理论、方法];
学科分类号
081202 ;
摘要
Video anomaly detection (VAD) is an important but challenging task in computer vision. The main challenge rises due to the rarity of training samples to model all anomaly cases. Hence, semi-supervised anomaly detection methods have gotten more attention, since they focus on modeling normals and they detect anomalies by measuring the deviations from normal patterns. Despite impressive advances of these methods in modeling normal motion and appearance, long-term motion modeling has not been effectively explored so far. Inspired by the abilities of the future frame prediction proxy-task, we introduce the task of future video prediction from a single frame, as a novel proxy-task for video anomaly detection. This proxy-task alleviates the challenges of previous methods in learning longer motion patterns. Moreover, we replace the initial and future raw frames with their corresponding semantic segmentation map, which not only makes the method aware of object class but also makes the prediction task less complex for the model. Extensive experiments on the benchmark datasets (ShanghaiTech, UCSD-Ped1, and UCSD-Ped2) show the effectiveness of the method and the superiority of its performance compared to SOTA prediction-based VAD methods.
引用
收藏
页码:472 / 486
页数:15
相关论文
共 50 条
  • [1] Future Frame Prediction Network for Video Anomaly Detection
    Luo, Weixin
    Liu, Wen
    Lian, Dongze
    Gao, Shenghua
    IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2022, 44 (11) : 7505 - 7520
  • [2] Video anomaly detection method based on future frame prediction and attention mechanism
    Wang, Chenxu
    Yao, Yanxin
    Yao, Han
    2021 IEEE 11TH ANNUAL COMPUTING AND COMMUNICATION WORKSHOP AND CONFERENCE (CCWC), 2021, : 405 - 407
  • [3] Robust Unsupervised Video Anomaly Detection by Multipath Frame Prediction
    Wang, Xuanzhao
    Che, Zhengping
    Jiang, Bo
    Xiao, Ning
    Yang, Ke
    Tang, Jian
    Ye, Jieping
    Wang, Jingyu
    Qi, Qi
    IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, 2022, 33 (06) : 2301 - 2312
  • [4] Motion-aware future frame prediction for video anomaly detection based on saliency perception
    Xu, Haitao
    Liu, Weibin
    Xing, Weiwei
    Wei, Xiang
    SIGNAL IMAGE AND VIDEO PROCESSING, 2022, 16 (08) : 2121 - 2129
  • [5] Motion-aware future frame prediction for video anomaly detection based on saliency perception
    Haitao Xu
    Weibin Liu
    Weiwei Xing
    Xiang Wei
    Signal, Image and Video Processing, 2022, 16 : 2121 - 2129
  • [6] SURVEILLANCE VIDEO ANOMALY DETECTION WITH FEATURE ENHANCEMENT AND CONSISTENCY FRAME PREDICTION
    Zou, Beiji
    Wang, Min
    Jiang, LingZi
    Zhang, Yue
    Liu, Shu
    2022 IEEE INTERNATIONAL CONFERENCE ON MULTIMEDIA AND EXPO WORKSHOPS (IEEE ICMEW 2022), 2022,
  • [7] Video Frame Prediction from a Single Image and Events
    Zhu, Juanjuan
    Wan, Zhexiong
    Dai, Yuchao
    THIRTY-EIGHTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, VOL 38 NO 7, 2024, : 7748 - 7756
  • [8] Video Anomaly Detection for Surveillance Based on Effective Frame Area
    Yang, Yuxing
    Xian, Yang
    Fu, Zeyu
    Naqvi, Syed Mohsen
    2021 IEEE 24TH INTERNATIONAL CONFERENCE ON INFORMATION FUSION (FUSION), 2021, : 1136 - 1140
  • [9] A Kalman Variational Autoencoder Model Assisted by Odometric Clustering for Video Frame Prediction and Anomaly Detection
    Slavic, Giulia
    Alemaw, Abrham Shiferaw
    Marcenaro, Lucio
    Gomez, David Martin
    Regazzoni, Carlo
    IEEE TRANSACTIONS ON IMAGE PROCESSING, 2023, 32 : 415 - 429
  • [10] Future Frame Prediction for Anomaly Detection - A New Baseline
    Liu, Wen
    Luo, Weixin
    Lian, Dongze
    Gao, Shenghua
    2018 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2018, : 6536 - 6545