Abstract
Contextual Importance and Utility (CIU) is a model-agnostic method for explaining outcomes of AI systems. CIU has succeeded in producing meaningful explanations where state-of-the-art methods fail, e.g. for detecting bleeding in gastroenterological images. This paper presents a Python implementation of CIU for explaining image classifications.
| Original language | English |
|---|---|
| Title of host publication | Explainable and Transparent AI and Multi-Agent Systems - 6th International Workshop, EXTRAAMAS 2024, Revised Selected Papers |
| Editors | Davide Calvaresi, Amro Najjar, Andrea Omicini, Rachele Carli, Giovanni Ciatto, Reyhan Aydogan, Joris Hulstijn, Kary Främling |
| Place of Publication | Cham |
| Publisher | Springer |
| Pages | 184-188 |
| Number of pages | 5 |
| ISBN (Electronic) | 978-3-031-70074-3 |
| ISBN (Print) | 978-3-031-70073-6 |
| DOIs | |
| Publication status | Published - 2024 |
| MoE publication type | A4 Conference publication |
| Event | International Workshop on Explainable, Transparent Autonomous Agents and Multi-Agent Systems - Auckland, New Zealand Duration: 6 May 2024 → 10 May 2024 Conference number: 6 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Publisher | Springer |
| Volume | 14847 LNAI |
| ISSN (Print) | 0302-9743 |
| ISSN (Electronic) | 1611-3349 |
Workshop
| Workshop | International Workshop on Explainable, Transparent Autonomous Agents and Multi-Agent Systems |
|---|---|
| Abbreviated title | EXTRAAMAS |
| Country/Territory | New Zealand |
| City | Auckland |
| Period | 06/05/2024 → 10/05/2024 |
Funding
The work is partially supported by the Wallenberg AI, Autonomous Systems and Software Program (WASP) funded by the Knut and Alice Wallenberg Foundation.
Keywords
- Contextual Importance and Utility
- Deep Neural Network
- Explainable Artificial Intelligence
- Image Classification
Fingerprint
Dive into the research topics of 'py_ciu_image : A Python Library for Explaining Image Classification with Contextual Importance and Utility'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver