Matches in SemOpenAlex for { <https://semopenalex.org/work/W4387415072> ?p ?o ?g. }
Showing items 1 to 70 of
70
with 100 items per page.
- W4387415072 endingPage "18" @default.
- W4387415072 startingPage "1" @default.
- W4387415072 abstract "Anomaly detection has recently gained increasing attention in the field of computer vision, likely due to its broad set of applications ranging from product fault detection on industrial production lines and impending event detection in video surveillance to finding lesions in medical scans. Regardless of the domain, anomaly detection is typically framed as a one-class classification task, where the learning is conducted on normal examples only. An entire family of successful anomaly detection methods is based on learning to reconstruct masked normal inputs (eg patches, future frames, etc.) and exerting the magnitude of the reconstruction error as an indicator for the abnormality level. Unlike other reconstruction-based methods, we present a novel self-supervised masked convolutional transformer block (SSMCTB) that comprises the reconstruction-based functionality at a core architectural level. The proposed self-supervised block is extremely flexible, enabling information masking at any layer of a neural network and being compatible with a wide range of neural architectures. In this work, we extend our previous self-supervised predictive convolutional attentive block (SSPCAB) with a 3D masked convolutional layer, a transformer for channel-wise attention, as well as a novel self-supervised objective based on Huber loss. Furthermore, we show that our block is applicable to a wider variety of tasks, adding anomaly detection in medical images and thermal videos to the previously considered tasks based on RGB images and surveillance videos. We exhibit the generality and flexibility of SSMCTB by integrating it into multiple state-of-the-art neural models for anomaly detection, bringing forth empirical results that confirm considerable performance improvements on five benchmarks: MVTec AD, BRATS, Avenue, ShanghaiTech, and Thermal Rare Event. We release our code and data as open source at: <uri xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>https://github.com/ristea/ssmctb</uri> ." @default.
- W4387415072 created "2023-10-07" @default.
- W4387415072 creator A5022176859 @default.
- W4387415072 creator A5041199606 @default.
- W4387415072 creator A5079112987 @default.
- W4387415072 creator A5080823547 @default.
- W4387415072 creator A5081017623 @default.
- W4387415072 creator A5083932459 @default.
- W4387415072 creator A5089591955 @default.
- W4387415072 date "2023-01-01" @default.
- W4387415072 modified "2023-10-16" @default.
- W4387415072 title "Self-Supervised Masked Convolutional Transformer Block for Anomaly Detection" @default.
- W4387415072 doi "https://doi.org/10.1109/tpami.2023.3322604" @default.
- W4387415072 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/37801379" @default.
- W4387415072 hasPublicationYear "2023" @default.
- W4387415072 type Work @default.
- W4387415072 citedByCount "0" @default.
- W4387415072 crossrefType "journal-article" @default.
- W4387415072 hasAuthorship W4387415072A5022176859 @default.
- W4387415072 hasAuthorship W4387415072A5041199606 @default.
- W4387415072 hasAuthorship W4387415072A5079112987 @default.
- W4387415072 hasAuthorship W4387415072A5080823547 @default.
- W4387415072 hasAuthorship W4387415072A5081017623 @default.
- W4387415072 hasAuthorship W4387415072A5083932459 @default.
- W4387415072 hasAuthorship W4387415072A5089591955 @default.
- W4387415072 hasConcept C108583219 @default.
- W4387415072 hasConcept C119857082 @default.
- W4387415072 hasConcept C136389625 @default.
- W4387415072 hasConcept C153180895 @default.
- W4387415072 hasConcept C154945302 @default.
- W4387415072 hasConcept C2524010 @default.
- W4387415072 hasConcept C2777210771 @default.
- W4387415072 hasConcept C31972630 @default.
- W4387415072 hasConcept C33923547 @default.
- W4387415072 hasConcept C41008148 @default.
- W4387415072 hasConcept C50644808 @default.
- W4387415072 hasConcept C739882 @default.
- W4387415072 hasConcept C81363708 @default.
- W4387415072 hasConceptScore W4387415072C108583219 @default.
- W4387415072 hasConceptScore W4387415072C119857082 @default.
- W4387415072 hasConceptScore W4387415072C136389625 @default.
- W4387415072 hasConceptScore W4387415072C153180895 @default.
- W4387415072 hasConceptScore W4387415072C154945302 @default.
- W4387415072 hasConceptScore W4387415072C2524010 @default.
- W4387415072 hasConceptScore W4387415072C2777210771 @default.
- W4387415072 hasConceptScore W4387415072C31972630 @default.
- W4387415072 hasConceptScore W4387415072C33923547 @default.
- W4387415072 hasConceptScore W4387415072C41008148 @default.
- W4387415072 hasConceptScore W4387415072C50644808 @default.
- W4387415072 hasConceptScore W4387415072C739882 @default.
- W4387415072 hasConceptScore W4387415072C81363708 @default.
- W4387415072 hasLocation W43874150721 @default.
- W4387415072 hasLocation W43874150722 @default.
- W4387415072 hasOpenAccess W4387415072 @default.
- W4387415072 hasPrimaryLocation W43874150721 @default.
- W4387415072 hasRelatedWork W2611989081 @default.
- W4387415072 hasRelatedWork W3029198973 @default.
- W4387415072 hasRelatedWork W3133861977 @default.
- W4387415072 hasRelatedWork W3167935049 @default.
- W4387415072 hasRelatedWork W3193565141 @default.
- W4387415072 hasRelatedWork W4226493464 @default.
- W4387415072 hasRelatedWork W4293226380 @default.
- W4387415072 hasRelatedWork W4312417841 @default.
- W4387415072 hasRelatedWork W4313906399 @default.
- W4387415072 hasRelatedWork W4375867731 @default.
- W4387415072 isParatext "false" @default.
- W4387415072 isRetracted "false" @default.
- W4387415072 workType "article" @default.