Weakly-supervised video anomaly detection with contrastive learning of long and short-range temporal features
Anomaly detection with weakly supervised video-level labels is typically formulated as a multiple instance learning (MIL) problem, in which we aim to identify snippets containing abnormal events, with each video represented as a bag of video snippets. Although current methods show effective detectio...
Saved in:
Main Authors: | , , , , , |
---|---|
Format: | text |
Language: | English |
Published: |
Institutional Knowledge at Singapore Management University
2021
|
Subjects: | |
Online Access: | https://ink.library.smu.edu.sg/sis_research/7021 https://ink.library.smu.edu.sg/context/sis_research/article/8024/viewcontent/2101.10030.pdf |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Singapore Management University |
Language: | English |
id |
sg-smu-ink.sis_research-8024 |
---|---|
record_format |
dspace |
spelling |
sg-smu-ink.sis_research-80242022-09-22T07:55:36Z Weakly-supervised video anomaly detection with contrastive learning of long and short-range temporal features TIAN, Yu PANG, Guansong CHEN, Yuanhong SINGH, Rajvinder VERJANS, Johan W. CARNEIRO, Gustavo Anomaly detection with weakly supervised video-level labels is typically formulated as a multiple instance learning (MIL) problem, in which we aim to identify snippets containing abnormal events, with each video represented as a bag of video snippets. Although current methods show effective detection performance, their recognition of the positive instances, i.e., rare abnormal snippets in the abnormal videos, is largely biased by the dominant negative instances, especially when the abnormal events are subtle anomalies that exhibit only small differences compared with normal events. This issue is exacerbated in many methods that ignore important video temporal dependencies. To address this issue, we introduce a novel and theoretically sound method, named Robust Temporal Feature Magnitude learning (RTFM), which trains a feature magnitude learning function to effectively recognise the positive instances, substantially improving the robustness of the MIL approach to the negative instances from abnormal videos. RTFM also adapts dilated convolutions and self-attention mechanisms to capture long- and short-range temporal dependencies to learn the feature magnitude more faithfully. Extensive experiments show that the RTFM-enabled MIL model (i) outperforms several state-of-the-art methods by a large margin on four benchmark data sets (ShanghaiTech, UCF-Crime, XD-Violence and UCSD-Peds) and (ii) achieves significantly improved subtle anomaly discriminability and sample efficiency. 2021-10-01T07:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/7021 info:doi/10.1109/ICCV48922.2021.00493 https://ink.library.smu.edu.sg/context/sis_research/article/8024/viewcontent/2101.10030.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ Research Collection School Of Computing and Information Systems eng Institutional Knowledge at Singapore Management University Vision applications and systems Action and behavior recognition Transfer/Low-shot/Semi/Unsupervised Learning Artificial Intelligence and Robotics Databases and Information Systems |
institution |
Singapore Management University |
building |
SMU Libraries |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
SMU Libraries |
collection |
InK@SMU |
language |
English |
topic |
Vision applications and systems Action and behavior recognition Transfer/Low-shot/Semi/Unsupervised Learning Artificial Intelligence and Robotics Databases and Information Systems |
spellingShingle |
Vision applications and systems Action and behavior recognition Transfer/Low-shot/Semi/Unsupervised Learning Artificial Intelligence and Robotics Databases and Information Systems TIAN, Yu PANG, Guansong CHEN, Yuanhong SINGH, Rajvinder VERJANS, Johan W. CARNEIRO, Gustavo Weakly-supervised video anomaly detection with contrastive learning of long and short-range temporal features |
description |
Anomaly detection with weakly supervised video-level labels is typically formulated as a multiple instance learning (MIL) problem, in which we aim to identify snippets containing abnormal events, with each video represented as a bag of video snippets. Although current methods show effective detection performance, their recognition of the positive instances, i.e., rare abnormal snippets in the abnormal videos, is largely biased by the dominant negative instances, especially when the abnormal events are subtle anomalies that exhibit only small differences compared with normal events. This issue is exacerbated in many methods that ignore important video temporal dependencies. To address this issue, we introduce a novel and theoretically sound method, named Robust Temporal Feature Magnitude learning (RTFM), which trains a feature magnitude learning function to effectively recognise the positive instances, substantially improving the robustness of the MIL approach to the negative instances from abnormal videos. RTFM also adapts dilated convolutions and self-attention mechanisms to capture long- and short-range temporal dependencies to learn the feature magnitude more faithfully. Extensive experiments show that the RTFM-enabled MIL model (i) outperforms several state-of-the-art methods by a large margin on four benchmark data sets (ShanghaiTech, UCF-Crime, XD-Violence and UCSD-Peds) and (ii) achieves significantly improved subtle anomaly discriminability and sample efficiency. |
format |
text |
author |
TIAN, Yu PANG, Guansong CHEN, Yuanhong SINGH, Rajvinder VERJANS, Johan W. CARNEIRO, Gustavo |
author_facet |
TIAN, Yu PANG, Guansong CHEN, Yuanhong SINGH, Rajvinder VERJANS, Johan W. CARNEIRO, Gustavo |
author_sort |
TIAN, Yu |
title |
Weakly-supervised video anomaly detection with contrastive learning of long and short-range temporal features |
title_short |
Weakly-supervised video anomaly detection with contrastive learning of long and short-range temporal features |
title_full |
Weakly-supervised video anomaly detection with contrastive learning of long and short-range temporal features |
title_fullStr |
Weakly-supervised video anomaly detection with contrastive learning of long and short-range temporal features |
title_full_unstemmed |
Weakly-supervised video anomaly detection with contrastive learning of long and short-range temporal features |
title_sort |
weakly-supervised video anomaly detection with contrastive learning of long and short-range temporal features |
publisher |
Institutional Knowledge at Singapore Management University |
publishDate |
2021 |
url |
https://ink.library.smu.edu.sg/sis_research/7021 https://ink.library.smu.edu.sg/context/sis_research/article/8024/viewcontent/2101.10030.pdf |
_version_ |
1770576189367255040 |