lukas/bachelor-thesis
1
0
Fork 0
Code Issues Pull Requests Actions Packages Releases Activity

improve intro
All checks were successful
Build Typst document / build_typst_documents (push) Successful in 3m26s
Details

Browse Source
This commit is contained in:
lukas-heiligenbrunner 2024-11-29 16:18:04 +01:00
parent 22e1edf077
commit a3ba4cc30b
2 changed files with 6 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
introduction.typ
View File

@ -3,9 +3,12 @@
Anomaly detection has especially in the industrial and automotive field essential importance. Anomaly detection has especially in the industrial and automotive field essential importance.
Lots of assembly lines need visual inspection to find errors often with the help of camera systems. Lots of assembly lines need visual inspection to find errors often with the help of camera systems.
Machine learning helped the field to advance a lot in the past. Machine learning helped the field to advance a lot in the past.
Most of the time the error rate is sub $.1%$ and therefore plenty of good data is available and the data is heavily unbalaned.
PatchCore and EfficientAD are state of the art algorithms trained only on good data and then detect anomalies within unseen (but similar) data. PatchCore and EfficientAD are state of the art algorithms trained only on good data and then detect anomalies within unseen (but similar) data.
One of their problems is the need of lots of training data and time to train. One of their problems is the need of lots of training data and time to train.
Few-Shot learning might be a suitable alternative with essentially lowered train time. Moreover a slight change of the camera position or the lighting conditions can lead to a complete retraining of the model.
Few-Shot learning might be a suitable alternative with hugely lowered train times and fast adaption to new conditions.
In this thesis the performance of 3 Few-Shot learning algorithms will be compared in the field of anomaly detection. In this thesis the performance of 3 Few-Shot learning algorithms will be compared in the field of anomaly detection.
Moreover, few-shot learning might be able not only to detect anomalies but also to detect the anomaly class. Moreover, few-shot learning might be able not only to detect anomalies but also to detect the anomaly class.

4
materialandmethods.typ
View File

@ -16,7 +16,7 @@ Each category comprises a set of defect-free training images and a test set of i
In this bachelor thesis only two categories are used. The categories are "Bottle" and "Cable". In this bachelor thesis only two categories are used. The categories are "Bottle" and "Cable".
The bottle category contains of 3 different defect classes: 'broken_large', 'broken_small' and 'contamination'. The bottle category contains 3 different defect classes: 'broken_large', 'broken_small' and 'contamination'.
#subpar.grid( #subpar.grid(
figure(image("rsc/mvtec/bottle/broken_large_example.png"), caption: [ figure(image("rsc/mvtec/bottle/broken_large_example.png"), caption: [
Broken large defect Broken large defect
@ -34,7 +34,7 @@ The bottle category contains of 3 different defect classes: 'broken_large', 'bro
Whereas cable has a lot more defect classes: 'bent_wire', 'cable_swap', 'combined', 'cut_inner_insulation', Whereas cable has a lot more defect classes: 'bent_wire', 'cable_swap', 'combined', 'cut_inner_insulation',
'cut_outer_insulation', 'missing_cable', 'missing_wire', 'poke_insulation'. 'cut_outer_insulation', 'missing_cable', 'missing_wire', 'poke_insulation'.
So many more defect classes are already an indication that a classification task will be more difficult for the cable category. So many more defect classes are already an indication that a classification task might be more difficult for the cable category.
#subpar.grid( #subpar.grid(
figure(image("rsc/mvtec/cable/bent_wire_example.png"), caption: [ figure(image("rsc/mvtec/cable/bent_wire_example.png"), caption: [