Header

UZH-Logo

Maintenance Infos

Selection of XAI Methods Matters: Evaluation of Feature Attribution Methods for Oculomotoric Biometric Identification


Krakowczyk, Daniel; Reich, David R; Prasse, Paul; Lapuschkin, Sebastian; Scheffer, Tobias; Jäger, Lena A (2022). Selection of XAI Methods Matters: Evaluation of Feature Attribution Methods for Oculomotoric Biometric Identification. In: NeurIPS 2022, Gaze Meets ML Workshop, New Orleans, LA, 3 December 2022, 1-29.

Abstract

Substantial advances in oculomotoric biometric identification have been made due to deep neural networks processing non-aggregated time series data that replace methods processing theoretically motivated engineered features. However, interpretability of deep neural networks is not trivial and needs to be thoroughly investigated for future eye tracking applications. Especially in medical or legal applications explanations can be required to be provided alongside predictions. In this work, we apply several attribution methods to a state of the art model for eye movement-based biometric identification. To asses the quality of the generated attributions, this work is focused on the quantitative evaluation of a range of established metrics. We find that Layer-wise Relevance Propagation generates the least complex attributions, while DeepLIFT attributions are the most faithful. Due to the absence of a correlation between attributions of these two methods we advocate to consider both methods for their potentially complementary attributions.

Abstract

Substantial advances in oculomotoric biometric identification have been made due to deep neural networks processing non-aggregated time series data that replace methods processing theoretically motivated engineered features. However, interpretability of deep neural networks is not trivial and needs to be thoroughly investigated for future eye tracking applications. Especially in medical or legal applications explanations can be required to be provided alongside predictions. In this work, we apply several attribution methods to a state of the art model for eye movement-based biometric identification. To asses the quality of the generated attributions, this work is focused on the quantitative evaluation of a range of established metrics. We find that Layer-wise Relevance Propagation generates the least complex attributions, while DeepLIFT attributions are the most faithful. Due to the absence of a correlation between attributions of these two methods we advocate to consider both methods for their potentially complementary attributions.

Statistics

Downloads

24 downloads since deposited on 13 Feb 2023
2 downloads since 12 months
Detailed statistics

Additional indexing

Item Type:Conference or Workshop Item (Paper), refereed, original work
Communities & Collections:06 Faculty of Arts > Department of Comparative Language Science
06 Faculty of Arts > Institute of Computational Linguistics
08 Research Priority Programs > Digital Society Initiative
06 Faculty of Arts > Zurich Center for Linguistics
Dewey Decimal Classification:000 Computer science, knowledge & systems
600 Technology
620 Engineering
Uncontrolled Keywords:eye-tracking, eye movements, biometrics, explainable artificial intelligence, XAI
Language:English
Event End Date:3 December 2022
Deposited On:13 Feb 2023 17:21
Last Modified:23 Mar 2023 10:25
OA Status:Green
Free access at:Official URL. An embargo period may apply.
Official URL:https://openreview.net/pdf?id=GOLdDAP2AtI
  • Content: Published Version
  • Language: English