| Autor: |
Kirchknopf, Armin, Slijepcevic, Djordje, Wunderlich, Ilkay, Breiter, Michael, Traxler, Johannes, Zeppelzauer, Matthias |
| Rok vydání: |
2022 |
| Předmět: |
|
| Zdroj: |
Proceedings of the Workshop of the Austrian Association for Pattern Recognition 2021 |
| Druh dokumentu: |
Working Paper |
| DOI: |
10.3217/978-3-85125-869-1-13 |
| Popis: |
We investigate the problem of explainability for visual object detectors. Specifically, we demonstrate on the example of the YOLO object detector how to integrate Grad-CAM into the model architecture and analyze the results. We show how to compute attribution-based explanations for individual detections and find that the normalization of the results has a great impact on their interpretation. |
| Databáze: |
arXiv |
| Externí odkaz: |
|
