Details
| Originalsprache | Englisch |
|---|---|
| Titel des Sammelwerks | Interspeech 2022 |
| Untertitel | Incheon, Korea 18-22 September 2022 |
| Seiten | 4003-4007 |
| Seitenumfang | 5 |
| Publikationsstatus | Veröffentlicht - 2022 |
| Veranstaltung | 23rd Annual Conference of the International Speech Communication Association, INTERSPEECH 2022 - Incheon, Südkorea Dauer: 18 Sept. 2022 → 22 Sept. 2022 |
Publikationsreihe
| Name | Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH |
|---|---|
| ISSN (Print) | 2308-457X |
Abstract
Respiratory sound classification is an important tool for remote screening of respiratory-related diseases such as pneumonia, asthma, and COVID-19. To facilitate the interpretability of classification results, especially ones based on deep learning, many explanation methods have been proposed using prototypes. However, existing explanation techniques often assume that the data is non-biased and the prediction results can be explained by a set of prototypical examples. In this work, we develop a unified example-based explanation method for selecting both representative data (prototypes) and outliers (criticisms). In particular, we propose a novel application of adversarial attacks to generate an explanation spectrum of data instances via an iterative fast gradient sign method. Such unified explanation can avoid over-generalisation and bias by allowing human experts to assess the model mistakes case by case. We performed a wide range of quantitative and qualitative evaluations to show that our approach generates effective and understandable explanation and is robust with many deep learning models.
ASJC Scopus Sachgebiete
- Geisteswissenschaftliche Fächer (insg.)
- Sprache und Linguistik
- Informatik (insg.)
- Mensch-Maschine-Interaktion
- Informatik (insg.)
- Signalverarbeitung
- Informatik (insg.)
- Software
- Mathematik (insg.)
- Modellierung und Simulation
Ziele für nachhaltige Entwicklung
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
Interspeech 2022 : Incheon, Korea 18-22 September 2022. 2022. S. 4003-4007 (Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH).
Publikation: Beitrag in Buch/Bericht/Sammelwerk/Konferenzband › Aufsatz in Konferenzband › Forschung › Peer-Review
}
TY - GEN
T1 - Example-based Explanations with Adversarial Attacks for Respiratory Sound Analysis
AU - Chang, Yi
AU - Ren, Zhao
AU - Nguyen, Thanh Tam
AU - Nejdl, Wolfgang
AU - Schuller, Björn W.
N1 - Funding Information: * Y. Chang and Z. Ren contribute equally. T. T. Nguyen is the corresponding author. This research was partially funded by the Federal Ministry of Education and Research (BMBF), Germany under the project LeibnizKILabor with grant No. 01DD20003 and the research projects “IIP-Ecosphere”, granted by the German Federal Ministry for Economics and Climate Action (BMWK) via funding code No. 01MK20006A. The code is released at https://github.com/ glam-imperial/SoundPrototypeCriticism.
PY - 2022
Y1 - 2022
N2 - Respiratory sound classification is an important tool for remote screening of respiratory-related diseases such as pneumonia, asthma, and COVID-19. To facilitate the interpretability of classification results, especially ones based on deep learning, many explanation methods have been proposed using prototypes. However, existing explanation techniques often assume that the data is non-biased and the prediction results can be explained by a set of prototypical examples. In this work, we develop a unified example-based explanation method for selecting both representative data (prototypes) and outliers (criticisms). In particular, we propose a novel application of adversarial attacks to generate an explanation spectrum of data instances via an iterative fast gradient sign method. Such unified explanation can avoid over-generalisation and bias by allowing human experts to assess the model mistakes case by case. We performed a wide range of quantitative and qualitative evaluations to show that our approach generates effective and understandable explanation and is robust with many deep learning models.
AB - Respiratory sound classification is an important tool for remote screening of respiratory-related diseases such as pneumonia, asthma, and COVID-19. To facilitate the interpretability of classification results, especially ones based on deep learning, many explanation methods have been proposed using prototypes. However, existing explanation techniques often assume that the data is non-biased and the prediction results can be explained by a set of prototypical examples. In this work, we develop a unified example-based explanation method for selecting both representative data (prototypes) and outliers (criticisms). In particular, we propose a novel application of adversarial attacks to generate an explanation spectrum of data instances via an iterative fast gradient sign method. Such unified explanation can avoid over-generalisation and bias by allowing human experts to assess the model mistakes case by case. We performed a wide range of quantitative and qualitative evaluations to show that our approach generates effective and understandable explanation and is robust with many deep learning models.
KW - explainable machine learning
KW - interpretable methods
KW - Respiratory sound analysis
UR - http://www.scopus.com/inward/record.url?scp=85140094241&partnerID=8YFLogxK
U2 - https://doi.org/10.48550/arXiv.2203.16141
DO - https://doi.org/10.48550/arXiv.2203.16141
M3 - Conference contribution
AN - SCOPUS:85140094241
T3 - Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH
SP - 4003
EP - 4007
BT - Interspeech 2022
T2 - 23rd Annual Conference of the International Speech Communication Association, INTERSPEECH 2022
Y2 - 18 September 2022 through 22 September 2022
ER -