SIDU: Similarity Difference And Uniqueness Method for Explainable AI

Muddamsetty, Satya M.; Mohammad, Norasiah; Moeslund, Thomas B.

doi:10.1109/icip40778.2020.9190952

Cited by 14 publications

(14 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Note that Grad-CAM is the generalization of CAM [870]. There are also some improved versions of Grad-CAM: Chattopadhay et al [103] propose a method that gives better object localization and improved heatmaps, when multiple object instances occur and in [501] a method with a more precise heatmap is described using a similarity difference mask, which is combined with a heatmap to generate the final heatmap. In addition, Wang et al [755] introduced a class activation mapping method, that removed the dependency on the gradients by using a linear combination of weights and activation maps.…”

Section: Model-specific Methodsmentioning

confidence: 99%

Knowledge Augmented Machine Learning with Applications in Autonomous Driving: A Survey

Wörmann¹,

Bogdoll²,

Bührle³

et al. 2022

Preprint

View full text Add to dashboard Cite

The existence of representative datasets is a prerequisite of many successful artificial intelligence and machine learning models. However, the subsequent application of these models often involves scenarios that are inadequately represented in the data used for training. The reasons for this are manifold and range from time and cost constraints to ethical considerations. As a consequence, the reliable use of these models, especially in safety-critical applications, is a huge challenge. Leveraging additional, already existing sources of knowledge is key to overcome the limitations of purely data-driven approaches, and eventually to increase the generalization capability of these models. Furthermore, predictions that conform with knowledge are crucial for making trustworthy and safe decisions even in underrepresented scenarios. This work provides an overview of existing techniques and methods in the literature that combine data-based models with existing knowledge. The identified approaches are structured according to the categories integration, extraction and conformity. Special attention is given to applications in the field of autonomous driving.Acknowledgement: The research leading to these results is funded by the German Federal Ministry for Economic Affairs and Climate Action within the project "KI Wissen -Entwicklung von Methoden für die Einbindung von Wissen in maschinelles Lernen". The authors would like to thank the consortium for the successful cooperation.

show abstract

Section: Model-specific Methodsmentioning

confidence: 99%

Knowledge Augmented Machine Learning with Applications in Autonomous Driving: A Survey

Wörmann¹,

Bogdoll²,

Bührle³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…A new visual explanation method known as SIDU proposed recently in [17] estimates the pixel saliency by extracting the last convolutional layer of the deep CNN model and creating the similarity difference mask which is eventually combined to form a final map for generating the visual explanation of the prediction. This method generates a heatmap based on two steps: Similarity difference and Uniqueness.…”

Section: Sidumentioning

confidence: 99%

“…Finally the heatmaps obtained via the eye-tracker can be compared directly by XAI methods. In this work we use heatmaps generated together by two XAI state-of-the-art methods namely SIDU [17], GRAD-CAM [23] using two different evaluation metrics.…”

Section: Introductionmentioning

confidence: 99%

Expert Level Evaluations for Explainable AI (XAI) Methods in the Medical Domain

Muddamsetty

Jahromi

Moeslund

2021

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

The recently emerged field of explainable artificial intelligence (XAI) attempts to shed lights on 'black box' Machine Learning (ML) models in understandable terms for human. As several explanation methods are developed alongside different applications for a black box model, the need for expert-level evaluation in inspecting their effectiveness becomes inevitable. This is significantly important for sensitive domains such as medical applications where evaluation of experts is essential to better understand how accurate the results of complex ML are and debug the models if necessary. The aim of this study is to experimentally show how the expert-level evaluation of XAI methods in a medical application can be utilized and aligned with the actual explanations generated by the clinician. To this end, we collect annotations from expert subjects equipped with an eye-tracker while they classify medical images and devise an approach for comparing the results with those obtained from XAI methods. We demonstrate the effectiveness of our approach in several experiments.

show abstract

“…This obviates the need to train interpretable classifiers for explaining each input-output relation (as in LIME) for every test point. Our proposed SIDU [13] method falls under perturbation-based methods but can effectively localize entire salient region of the object of interest compared to the state-of-the-art XAI methods such as Grad-CAM and RISE. Furthermore, it is less computationally complex.…”

Section: Visual Explanationmentioning

confidence: 99%

“…While each of these methods can be justifiable in one way or another, apart from challenges such as gradient computation of DNN architecture (e.g., Grad-CAM) or visualizing all the perturbations modes (e.g., RISE), the generated visual explanation suffers from a lack of localizing the entire salient regions of an object, which is often required for higher classification scores. To address this high-impact problem, we proposed a new visual explanation approach known as SIDU [13] for estimating pixel saliency by extracting the last convolutional layer of the deep CNN model and creating the similarity differences and uniqueness masks which are eventually combined to form a final map for generating the visual explanation for the prediction.…”

Section: Introductionmentioning

confidence: 99%

Introducing and assessing the explainable AI (XAI)method: SIDU

Muddamsetty¹,

Jahromi²,

Ciontos³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Explainable Artificial Intelligence (XAI) has in recent years become a well-suited framework to generate human understandable explanations of black box models. In this paper, we present a novel XAI visual explanation algorithm denoted SIDU that can effectively localize entire object regions responsible for prediction in a full extend. We analyze its robustness and effectiveness through various computational and human subject experiments. In particular, we assess the SIDU algorithm using three different types of evaluations (Application, Human and Functionally-Grounded) to demonstrate its superior performance. The robustness of SIDU is further studied in presence of adversarial attack on black box models to better understand its performance.

show abstract

SIDU: Similarity Difference And Uniqueness Method for Explainable AI

Cited by 14 publications

References 18 publications

Knowledge Augmented Machine Learning with Applications in Autonomous Driving: A Survey

Knowledge Augmented Machine Learning with Applications in Autonomous Driving: A Survey

Expert Level Evaluations for Explainable AI (XAI) Methods in the Medical Domain

Introducing and assessing the explainable AI (XAI)method: SIDU

Contact Info

Product

Resources

About