Industrial Anomaly Detection with Normalizing Flows

Marco Rudolph

doi:10.51202/9783186886101

Keywords

Search publication

Bibliographic data

Publication year: 2024
Publication date: 18.06.2024
ISBN-Print: 978-3-18-388610-4
ISBN-Online: 978-3-18-688610-1
Publisher: VDI Verlag, Düsseldorf
Series: Informatik/ Kommunikation
Volume: 886
Language: German
Pages: 132
Product type: Book Titles

Titelei/Inhaltsverzeichnis No access Pages I - XIV
1. Anomaly Detection in the Industrial Context No access
2. AddressedApplications No access
3. StructureoftheThesis No access
4. 1. Anomaly Detection No access
  2. OtherPublications No access
1. 1. Semi-Supervised Anomaly Detection No access
  2. Taxonomy No access
2. 1. Overview No access
  2. Traditional Approaches No access
  3. GenerativeModels No access
  4. Student-Teacher Networks No access
  5. DensityEstimation No access
  6. Synthetic Data No access
  7. Methods for Multivariate Machine Data No access
1. MVTecAD No access
2. MVTec 3D-AD No access
3. MagneticTileDefects No access
4. Metrics No access
1. 1. Introduction No access
  2. RelatedWork No access
  3. Real-NVP No access
2. 1. Architecture No access
  2. Training No access
  3. ScoringFunction No access
  4. Localization No access
3. 1. Implementation Details No access
  2. Detection No access
  3. Localization No access
  4. Ablation Studies No access
4. Conclusion No access
5. Open Problems No access
1. 1. Cross-ScaleFlow No access
  2. LearningObjective No access
  3. Localization No access
2. 1. Implementation Details No access
  2. Detection No access
  3. Localization No access
  4. Ablation Studies No access
3. Conclusion No access
4. Open Problems No access
1. 1. Teacher No access
  2. Student No access
  3. DPreprocessing No access
2. 1. Implementation Details No access
  2. Results No access
  3. Ablation Studies No access
3. Open Problems No access
4. Comparison with DifferNet and CS-Flow No access
5. Conclusion No access
1. Overview No access
2. Setup No access
3. RobotandSignals No access
4. Pick-And-Place Operation No access
5. 1. Process Errors No access
  2. Gripping Errors No access
  3. Robot axis wear No access
6. 1. Normal Data No access
  2. Anomalies No access
  3. Comparison to AURSAD No access
  4. Evaluation Protocol No access
7. Conclusion No access
1. 1. Architecture No access
  2. Training No access
  3. TemporalAnalysis No access
2. 1. Implementation Details No access
  2. Results No access
  3. Ablation Studies No access
  4. TemporalAnalysis No access
3. Conclusion No access
Conclusion No access Pages 110 - 114
Bibliography No access Pages 115 - 132