Unbiased multiple instance learning for weakly supervised video anomaly detection
Weakly Supervised Video Anomaly Detection (WSVAD) is challenging because the binary anomaly label is only given on the video level, but the output requires snippetlevel predictions. So, Multiple Instance Learning (MIL) is prevailing in WSVAD. However, MIL is notoriously known to suffer from man...
Saved in:
Main Authors: | , , , , , |
---|---|
Other Authors: | |
Format: | Conference or Workshop Item |
Language: | English |
Published: |
2023
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/169509 https://cvpr2023.thecvf.com/Conferences/2023 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
id |
sg-ntu-dr.10356-169509 |
---|---|
record_format |
dspace |
spelling |
sg-ntu-dr.10356-1695092023-08-29T00:58:03Z Unbiased multiple instance learning for weakly supervised video anomaly detection Lv, Hui Yue, Zhongqi Sun, Qianru Luo, Bin Cui, Zhen Zhang, Hanwang School of Computer Science and Engineering IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2023) Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence Multiple Instance Learning Weakly Supervised Video Anomaly Detection Weakly Supervised Video Anomaly Detection (WSVAD) is challenging because the binary anomaly label is only given on the video level, but the output requires snippetlevel predictions. So, Multiple Instance Learning (MIL) is prevailing in WSVAD. However, MIL is notoriously known to suffer from many false alarms because the snippet-level detector is easily biased towards the abnormal snippets with simple context, confused by the normality with the same bias, and missing the anomaly with a different pattern. To this end, we propose a new MIL framework: Unbiased MIL (UMIL), to learn unbiased anomaly features that improve WSVAD. At each MIL training iteration, we use the current detector to divide the samples into two groups with different context biases: the most confident abnormal/normal snippets and the rest ambiguous ones. Then, by seeking the invariant features across the two sample groups, we can remove the variant context biases. Extensive experiments on benchmarks UCF-Crime and TAD demonstrate the effectiveness of our UMIL. Our code is provided at https://github.com/ktr-hubrt/UMIL. Agency for Science, Technology and Research (A*STAR) AI Singapore Submitted/Accepted version The author gratefully acknowledges the support of Alibaba-NTU Singapore Joint Research Institute, the A*STAR under its AME YIRG Grant (Project No.A20E6c0101), the Lee Kong Chian (LKC) Fellowship fund awarded by Singapore Management University, AI Singapore AISG2-RP-2021-022, the Postgraduate Research & Practice Innovation Program of Jiangsu Province, the National Natural Science Foundation of China (Grants No.62072244), the Natural Science Foundation of Shandong Province (Grant No.ZR2020LZH008) and State Key Laboratory of High-end Server & Storage Technology. 2023-08-22T08:58:43Z 2023-08-22T08:58:43Z 2023 Conference Paper Lv, H., Yue, Z., Sun, Q., Luo, B., Cui, Z. & Zhang, H. (2023). Unbiased multiple instance learning for weakly supervised video anomaly detection. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2023). https://dx.doi.org/10.1109/CVPR52729.2023.00775 https://hdl.handle.net/10356/169509 10.1109/CVPR52729.2023.00775 https://cvpr2023.thecvf.com/Conferences/2023 en A20E6c0101 AISG2-RP-2021-022 © 2023 The Author(s). Published by Computer Vision Foundation. This is an open-access article distributed under the terms of the Creative Commons Attribution License. The final published version of the proceedings is available on IEEE Xplore. application/pdf |
institution |
Nanyang Technological University |
building |
NTU Library |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
NTU Library |
collection |
DR-NTU |
language |
English |
topic |
Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence Multiple Instance Learning Weakly Supervised Video Anomaly Detection |
spellingShingle |
Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence Multiple Instance Learning Weakly Supervised Video Anomaly Detection Lv, Hui Yue, Zhongqi Sun, Qianru Luo, Bin Cui, Zhen Zhang, Hanwang Unbiased multiple instance learning for weakly supervised video anomaly detection |
description |
Weakly Supervised Video Anomaly Detection (WSVAD)
is challenging because the binary anomaly label is only
given on the video level, but the output requires snippetlevel
predictions. So, Multiple Instance Learning (MIL) is
prevailing in WSVAD. However, MIL is notoriously known
to suffer from many false alarms because the snippet-level
detector is easily biased towards the abnormal snippets with
simple context, confused by the normality with the same
bias, and missing the anomaly with a different pattern. To
this end, we propose a new MIL framework: Unbiased MIL
(UMIL), to learn unbiased anomaly features that improve
WSVAD. At each MIL training iteration, we use the current
detector to divide the samples into two groups with different
context biases: the most confident abnormal/normal
snippets and the rest ambiguous ones. Then, by seeking
the invariant features across the two sample groups, we
can remove the variant context biases. Extensive experiments
on benchmarks UCF-Crime and TAD demonstrate
the effectiveness of our UMIL. Our code is provided at
https://github.com/ktr-hubrt/UMIL. |
author2 |
School of Computer Science and Engineering |
author_facet |
School of Computer Science and Engineering Lv, Hui Yue, Zhongqi Sun, Qianru Luo, Bin Cui, Zhen Zhang, Hanwang |
format |
Conference or Workshop Item |
author |
Lv, Hui Yue, Zhongqi Sun, Qianru Luo, Bin Cui, Zhen Zhang, Hanwang |
author_sort |
Lv, Hui |
title |
Unbiased multiple instance learning for weakly supervised video anomaly detection |
title_short |
Unbiased multiple instance learning for weakly supervised video anomaly detection |
title_full |
Unbiased multiple instance learning for weakly supervised video anomaly detection |
title_fullStr |
Unbiased multiple instance learning for weakly supervised video anomaly detection |
title_full_unstemmed |
Unbiased multiple instance learning for weakly supervised video anomaly detection |
title_sort |
unbiased multiple instance learning for weakly supervised video anomaly detection |
publishDate |
2023 |
url |
https://hdl.handle.net/10356/169509 https://cvpr2023.thecvf.com/Conferences/2023 |
_version_ |
1779156376613814272 |