Discrepancy Scaling for Fast Unsupervised Anomaly Localization

Forskningsoutput: Kapitel i bok/rapport/konferenshandlingKonferensbidragVetenskapligPeer review

Sammanfattning

Computer vision systems can automatically find and segment anomalies in images even without ever seeing anomalous observations during training. Many methods for such unsupervised anomaly detection (AD) and localization (AL) tasks have been introduced in recent years, but the most accurate methods tend to be computationally heavy. In this paper, we propose Discrepancy Scaling, a method that significantly improves the accuracy of a very fast AD and AL approach called Student-Teacher Feature Pyramid Matching. We show that with Discrepancy Scaling, even a small, mobile-friendly convolutional neural network can perform well on AD and AL tasks.
Originalspråkengelska
Titel på värdpublikation2023 IEEE 47th Annual Computers, Software, and Applications Conference (COMPSAC)
Antal sidor6
FörlagIEEE
Utgivningsdatumaug. 2023
Sidor274-279
ISBN (elektroniskt)979-8-3503-2697-0
DOI
StatusPublicerad - aug. 2023
MoE-publikationstypA4 Artikel i en konferenspublikation
EvenemangAnnual Computers, Software, and Applications Conference - Torino, Italien
Varaktighet: 26 juni 202330 juni 2023
Konferensnummer: 47

Vetenskapsgrenar

  • 113 Data- och informationsvetenskap

Citera det här