A Discrepancy Aware Framework for Robust Anomaly Detection

计算机科学 稳健性(进化) 异常检测 离群值 合成数据 边距(机器学习) 人工智能 机器学习 标记数据 编码(集合论) 解码方法 模式识别(心理学) 算法 生物化学 化学 集合(抽象数据类型) 基因 程序设计语言
作者
Yuxuan Cai,Dingkang Liang,Dongliang Luo,Xinwei He,Xin Yang,Xiang Bai
出处
期刊:Cornell University - arXiv
标识
DOI:10.48550/arxiv.2310.07585
摘要

Defect detection is a critical research area in artificial intelligence. Recently, synthetic data-based self-supervised learning has shown great potential on this task. Although many sophisticated synthesizing strategies exist, little research has been done to investigate the robustness of models when faced with different strategies. In this paper, we focus on this issue and find that existing methods are highly sensitive to them. To alleviate this issue, we present a Discrepancy Aware Framework (DAF), which demonstrates robust performance consistently with simple and cheap strategies across different anomaly detection benchmarks. We hypothesize that the high sensitivity to synthetic data of existing self-supervised methods arises from their heavy reliance on the visual appearance of synthetic data during decoding. In contrast, our method leverages an appearance-agnostic cue to guide the decoder in identifying defects, thereby alleviating its reliance on synthetic appearance. To this end, inspired by existing knowledge distillation methods, we employ a teacher-student network, which is trained based on synthesized outliers, to compute the discrepancy map as the cue. Extensive experiments on two challenging datasets prove the robustness of our method. Under the simple synthesis strategies, it outperforms existing methods by a large margin. Furthermore, it also achieves the state-of-the-art localization performance. Code is available at: https://github.com/caiyuxuan1120/DAF.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
更新
大幅提高文件上传限制,最高150M (2024-4-1)

科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
田様应助儒雅的夏山采纳,获得10
刚刚
1秒前
twb完成签到,获得积分10
1秒前
heheheli完成签到,获得积分10
2秒前
2秒前
传统的雨文完成签到,获得积分10
2秒前
ddl发布了新的文献求助10
3秒前
3秒前
davvero完成签到,获得积分10
3秒前
ldroc发布了新的文献求助10
4秒前
完美世界应助未来可期采纳,获得10
4秒前
4秒前
小秦同学发布了新的文献求助10
5秒前
6秒前
东方诩发布了新的文献求助10
6秒前
FRW发布了新的文献求助10
7秒前
Ava应助心灵美寄凡采纳,获得10
8秒前
8秒前
chouchou完成签到,获得积分10
10秒前
10秒前
彭于晏应助zjl123采纳,获得10
11秒前
科研宝完成签到,获得积分10
13秒前
Lucas应助昔时旧日采纳,获得10
13秒前
13秒前
14秒前
14秒前
思源应助怕黑妙之采纳,获得10
16秒前
(讼)发布了新的文献求助10
16秒前
想好好搞事业完成签到,获得积分10
16秒前
共享精神应助东方诩采纳,获得10
17秒前
烟花应助ddl采纳,获得10
19秒前
DrKe完成签到,获得积分10
19秒前
19秒前
slycmd完成签到,获得积分10
19秒前
20秒前
希望天下0贩的0应助CX330ren采纳,获得10
20秒前
白飞应助123采纳,获得10
21秒前
zjl123完成签到,获得积分10
22秒前
23秒前
24秒前
高分求助中
The late Devonian Standard Conodont Zonation 2000
Nickel superalloy market size, share, growth, trends, and forecast 2023-2030 2000
The Lali Section: An Excellent Reference Section for Upper - Devonian in South China 1500
Smart but Scattered: The Revolutionary Executive Skills Approach to Helping Kids Reach Their Potential (第二版) 1000
Very-high-order BVD Schemes Using β-variable THINC Method 850
Mantiden: Faszinierende Lauerjäger Faszinierende Lauerjäger 800
PraxisRatgeber: Mantiden: Faszinierende Lauerjäger 800
热门求助领域 (近24小时)
化学 医学 生物 材料科学 工程类 有机化学 生物化学 物理 内科学 纳米技术 计算机科学 化学工程 复合材料 基因 遗传学 催化作用 物理化学 免疫学 量子力学 细胞生物学
热门帖子
关注 科研通微信公众号,转发送积分 3251944
求助须知:如何正确求助?哪些是违规求助? 2894827
关于积分的说明 8283422
捐赠科研通 2563461
什么是DOI,文献DOI怎么找? 1391552
科研通“疑难数据库(出版商)”最低求助积分说明 651860
邀请新用户注册赠送积分活动 628894