Summary of On the Black-box Explainability Of Object Detection Models For Safe and Trustworthy Industrial Applications, by Alain Andres et al.
On the Black-box Explainability of Object Detection Models for Safe and Trustworthy Industrial Applications
by Alain Andres, Aitor Martinez-Seras, Ibai Laña, Javier Del Ser
First submitted to arxiv on: 28 Oct 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary A novel Explainable Artificial Intelligence (XAI) framework for object detection models is proposed in this work. The framework, dubbed D-MFPP, builds upon the Morphological Fragmental Perturbation Pyramid (MFPP) technique and utilizes segmentation-based masks to generate explanations. To evaluate the quality of these explanations, a new metric called D-Deletion is introduced, which combines faithfulness and localization to meet the unique demands of object detectors. The proposed methods are tested on real-world industrial and robotic datasets in two safety-critical environments: shared human-robot workspaces and assembly areas for battery kits. Results show that D-Deletion effectively gauges explanation performance when multiple elements of the same class appear, while D-MFPP provides a promising alternative to D-RISE with fewer masks. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about making artificial intelligence more understandable and safer to use in situations where mistakes could have serious consequences. It focuses on explaining why object detection models make certain decisions, like recognizing objects in pictures. The researchers developed two new methods for explaining these models: D-MFPP and D-Deletion. They tested these methods using real-world data from robots working together with humans or assembling battery components. This work can help create more reliable and trustworthy AI systems. |
Keywords
» Artificial intelligence » Object detection
