Filters








55,655 Hits in 3.9 sec

Mean-Shifted Contrastive Loss for Anomaly Detection [article]

Tal Reiss, Yedid Hoshen
2021 arXiv   pre-print
Our improvements yield a new anomaly detection approach, based on Mean-Shifted Contrastive Loss, which is both more accurate and less sensitive to catastrophic collapse than previous methods.  ...  In this paper, we propose a new loss function which can overcome failure modes of both center-loss and contrastive-loss methods.  ...  Our new loss function is named -Mean-Shifted Contrastive Loss.  ... 
arXiv:2106.03844v1 fatcat:iy3rrolgbjb5no562azav2tdxu

Pediatric Otoscopy Video Screening with Shift Contrastive Anomaly Detection [article]

Weiyao Wang, Aniruddha Tamhane, Christine Santos, John R. Rzasa, James H. Clark, Therese L. Canares, Mathias Unberath
2021 arXiv   pre-print
We present a two stage method that first, identifies valid frames by detecting and extracting ear drum patches from the video sequence, and second, performs the proposed shift contrastive anomaly detection  ...  Ear related concerns and symptoms represents the leading indication for seeking pediatric healthcare attention.  ...  Mean-Shifted Contrastive Loss: To adapt contrastive learning to anomaly detection, MSC [26] proposes the alternative mean-shifted contrastive loss.  ... 
arXiv:2110.13254v1 fatcat:b235bltgf5evdkug3xdvo2qx3i

Pediatric Otoscopy Video Screening With Shift Contrastive Anomaly Detection

Weiyao Wang, Aniruddha Tamhane, Christine Santos, John R. Rzasa, James H. Clark, Therese L. Canares, Mathias Unberath
2022 Frontiers in Digital Health  
We present a two stage method that first, identifies valid frames by detecting and extracting ear drum patches from the video sequence, and second, performs the proposed shift contrastive anomaly detection  ...  Ear related concerns and symptoms represent the leading indication for seeking pediatric healthcare attention.  ...  Mean-Shifted Contrastive Loss: To adapt contrastive learning to anomaly detection, MSC (27) proposes the alternative mean-shifted contrastive loss.  ... 
doi:10.3389/fdgth.2021.810427 pmid:35224535 pmcid:PMC8866874 fatcat:62dnalwr2basrh3bgqleja7qjq

Anomaly Detection in IR Images of PV Modules using Supervised Contrastive Learning [article]

Lukas Bommes, Mathis Hoffmann, Claudia Buerhop-Lutz, Tobias Pickel, Jens Hauch, Christoph Brabec, Andreas Maier, Ian Marius Peters
2021 arXiv   pre-print
We train a ResNet-34 convolutional neural network with a supervised contrastive loss, on top of which we employ a k-nearest neighbor classifier to detect anomalies.  ...  Our method is insensitive to hyperparameter settings, converges quickly and reliably detects unknown types of anomalies making it well suited for practice.  ...  ACKNOWLEDGEMENTS We would like to thank Sanjay Venugopal for valuable discussions about the contrastive learning objective.  ... 
arXiv:2112.02922v1 fatcat:clmfjwcy75gu7h2thekvixot2q

Self-Supervised Anomaly Detection: A Survey and Outlook [article]

Hadi Hojjati, Thi Kieu Khanh Ho, Narges Armanfard
2022 arXiv   pre-print
Finally, we discuss a variety of new directions for improving the existing algorithms.  ...  We also compare the performance of these models against each other and other state-of-the-art anomaly detection models.  ...  To overcome the hypersphere collapse problem, Reiss et al. (2021) [40] proposed a new loss function, called Mean-shifted contrastive loss (MSCL).  ... 
arXiv:2205.05173v2 fatcat:es7dkinhvrf7bepowfbbnj4hz4

Contrast vision and optic neuritis: neural blurring

R F Hess
1983 Journal of Neurology, Neurosurgery and Psychiatry  
Contrast detection and contrast matching experiments each demonstrate anomalies in contrast processing for eyes with optic neuritis.  ...  The mean luminance of each was equal and set to 1023  ...  Contrast detection and contrast matching experiments each demonstrate anomalies in contrast processing for eyes with optic neuritis.  ... 
doi:10.1136/jnnp.46.11.1023 pmid:6655476 pmcid:PMC491738 fatcat:z5y2xoj6h5hrdisji4jkiznx5i

Disentangling Physical Parameters for Anomalous Sound Detection Under Domain Shifts [article]

Kota Dohi, Takashi Endo, Yohei Kawaguchi
2021 arXiv   pre-print
To develop a sound-monitoring system for machines, a method for detecting anomalous sound under domain shifts is proposed. A domain shift occurs when a machine's physical parameters change.  ...  Anomaly scores calculated from this domain-shift-invariant latent space are unaffected by such shifts, which reduces false positives and improves the detection performance.  ...  Learning of domain-shift-invariant latent space for anomaly detection using NF To handle distribution changes due to domain shifts, we propose to disentangle the factors of domain shifts and construct  ... 
arXiv:2111.06539v1 fatcat:xskazn3jtnh5dcfsipzmw2m6du

Focus Your Distribution: Coarse-to-Fine Non-Contrastive Learning for Anomaly Detection and Localization [article]

Ye Zheng, Xiang Wang, Rui Deng, Tianpeng Bao, Rui Zhao, Liwei Wu
2021 arXiv   pre-print
Towards this end, we propose a novel framework for unsupervised anomaly detection and localization.  ...  The essence of unsupervised anomaly detection is to learn the compact distribution of normal samples and detect outliers as anomalies in testing.  ...  A natural algorithm for mode seeking is the mean-shift method proposed by (Comaniciu and Meer 2002) .  ... 
arXiv:2110.04538v1 fatcat:s5iueshbwvhi7noqaeldqkuskm

Self-supervised Anomaly Detection for New Physics [article]

Barry M. Dillon, Radha Mastandrea, Benjamin Nachman
2022 arXiv   pre-print
This opens the door to using low-dimensional latent representations as a computationally efficient space for resonant anomaly detection in generic particle collision events.  ...  We optimize the network using the self-supervised contrastive loss, which encourages the preservation of known physical symmetries of the dijets.  ...  However, additional augmentations for dijet events could be added to the contrastive loss.  ... 
arXiv:2205.10380v1 fatcat:xcyq6mnrr5chflshzuv36si4k4

Detecting network performance anomalies with contextual anomaly detection

Giorgos Dimopoulos, Pere Barlet-Ros, Constantine Dovrolis, Ilias Leontiadis
2017 2017 IEEE International Workshop on Measurement and Networking (M&N)  
To address this problem, we present in this paper a novel methodology for detecting performance anomalies based on contextual information.  ...  Network performance anomalies can be defined as abnormal and significant variations in a network's traffic levels. Being able to detect anomalies is critical for both network operators and end users.  ...  Fig. 9 : 9 Detection accuracy for level shift (a) and deviation shift (b) anomalies.  ... 
doi:10.1109/iwmn.2017.8078404 dblp:conf/iwmn/DimopoulosBDL17 fatcat:2zxvfxj7hjfxne3wczbvjzrigq

Latent Outlier Exposure for Anomaly Detection with Contaminated Data [article]

Chen Qiu, Aodong Li, Marius Kloft, Maja Rudolph, Stephan Mandt
2022 arXiv   pre-print
Inspired by outlier exposure (Hendrycks et al., 2018) that considers synthetically created, labeled anomalies, we thereby use a combination of two losses that share parameters: one for the normal and one  ...  We propose a strategy for training an anomaly detector in the presence of unlabeled anomalies that is compatible with a broad class of models.  ...  In contrast to most anomaly detection setups, we assume that our dataset is corrupted by anomalies.  ... 
arXiv:2202.08088v2 fatcat:3yowro6u7vannggi2jgdfngvwu

Thermal Anomalies Detect Critical Global Land Surface Changes

D. J. Mildrexler, M. Zhao, W. B. Cohen, S. W. Running, X. P. Song, M. O. Jones
2018 Journal of Applied Meteorology and Climatology  
The authors investigate whether maximum thermal anomalies from satellite observations could detect heat waves and droughts, a melting cryosphere, and disturbances in the tropical forest from 2003 to 2014  ...  sheets, severe droughts, and the incremental effects of forest loss in tropical forests.  ...  contrasting ecological conditions in these areas (Fig. 6) .  ... 
doi:10.1175/jamc-d-17-0093.1 fatcat:nqbytcbkbzdknpq4oeg7ozhv6m

Perceptual Image Anomaly Detection [article]

Nina Tuluptceva, Bart Bakker, Irina Fedulova, Anton Konushin
2019 arXiv   pre-print
We present a novel method for image anomaly detection, where algorithms that use samples drawn from some distribution of "normal" data, aim to detect out-of-distribution (abnormal) samples.  ...  It leverages Generative Adversarial Networks to learn these data distributions and uses perceptual loss for the detection of image abnormality.  ...  shifted by 5 pixels.  ... 
arXiv:1909.05904v1 fatcat:dlzcp7s6nvf3rndw3vvbnrerqq

Network-Wide Traffic Anomaly Detection and Localization Based on Robust Multivariate Probabilistic Calibration Model

Yuchong Li, Xingguo Luo, Yekui Qian, Xin Zhao
2015 Mathematical Problems in Engineering  
In order to solve the mentioned problems, this paper presents a robust multivariate probabilistic calibration model for network-wide anomaly detection and localization.  ...  However, when facing the actual problems of noise interference or data loss, the network-wide anomaly detection approaches also suffer significant performance reduction or may even become unavailable.  ...  Both methods succeed in detecting 6 anomalies, but the latter does not detect anomalies every time in the corresponding cycles of anomaly occurrence, especially for the 1200-1239 ingress/egress shifts.  ... 
doi:10.1155/2015/923792 fatcat:ve4ba6pfrvbzfnggt5gqnkolq4

Anomaly Detection in Autonomous Driving: A Survey [article]

Daniel Bogdoll, Maximilian Nitsche, J. Marius Zöllner
2022 arXiv   pre-print
We provide a systematization including detection approach, corner case level, ability for an online application, and further attributes.  ...  This survey provides an extensive overview of anomaly detection techniques based on camera, lidar, radar, multimodal and abstract object level data.  ...  Acknowledgment This work results from the project KI Data Tooling (19A20001J), funded by the German Federal Ministry for Economic Affairs and Climate Action (BMWK).  ... 
arXiv:2204.07974v1 fatcat:3rdola4tjfesllr6dqhqgf3tve
« Previous Showing results 1 — 15 out of 55,655 results