Analysis of Explainers of Black Box Deep Neural Networks for Computer Vision: A Survey

Buhrmester, Vanessa; Münch, David; Arens, Michael

doi:10.48550/arxiv.1911.12116

Cited by 28 publications

(36 citation statements)

References 62 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With this parameterization, we interpret the proportionality constant τ (k) 1 [m] as learning the gain factor between the image domain noise-level σ and the m-th subband's noise-level at layer k. This framework has the added benefit of decoupling noise-level estimation from denoising, allowing for trade-off between accurate estimation and speed at inference time. We explore this trade-off using two different noise-level estimation algorithms at inference time in Section IV-D.…”

Section: Noise-adaptive Thresholdsmentioning

confidence: 99%

CDLNet: Noise-Adaptive Convolutional Dictionary Learning Network for Blind Denoising and Demosaicing

Janjušević,

Khalilian-Gourtani,

Wang

2021

Preprint

View full text Add to dashboard Cite

Deep learning based methods hold state-of-the-art results in low-level image processing tasks, but remain difficult to interpret due to their black-box construction. Unrolled optimization networks present an interpretable alternative to constructing deep neural networks by deriving their architecture from classical iterative optimization methods without use of tricks from the standard deep learning tool-box. So far, such methods have demonstrated performance close to that of state-of-the-art models while using their interpretable construction to achieve a comparably low learned parameter count. In this work, we propose an unrolled convolutional dictionary learning network (CDLNet) and demonstrate its competitive denoising and joint denoising and demosaicing (JDD) performance both in low and high parameter count regimes. Specifically, we show that the proposed model outperforms state-of-the-art fully convolutional denoising and JDD models when scaled to a similar parameter count. In addition, we leverage the model's interpretable construction to propose a noise-adaptive parameterization of thresholds in the network that enables stateof-the-art blind denoising performance, and near-perfect generalization on noise-levels unseen during training. Furthermore, we show that such performance extends to the JDD task and unsupervised learning.

show abstract

Section: Noise-adaptive Thresholdsmentioning

confidence: 99%

CDLNet: Noise-Adaptive Convolutional Dictionary Learning Network for Blind Denoising and Demosaicing

Janjušević,

Khalilian-Gourtani,

Wang

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…The Explainable AI literature is blooming parallelly with the advances of DL models, and so is the set of surveys doing a great job at classifying the various methods [5,14,38]. We particularly focus on attribution methods, i.e.…”

Section: Related Work: Explainable Deep Learning and Compositional Pa...mentioning

confidence: 99%

EXplainable Neural-Symbolic Learning (X-NeSyL) methodology to fuse deep learning representations with expert knowledge graphs: the MonuMAI cultural heritage use case

Díaz-Rodríguez,

Lamas,

Sanchez

et al. 2021

Preprint

View full text Add to dashboard Cite

The latest Deep Learning (DL) models for detection and classification have achieved an unprecedented performance over classical machine learning algorithms. However, DL models are black-box methods hard to debug, interpret, and certify. DL alone cannot provide explanations that can be validated by a non technical audience such as end-users or domain experts. In contrast, symbolic AI systems that convert concepts into rules or symbols -such as knowledge graphs-are easier to explain. However, they present lower generalisation and scaling capabilities. A very important challenge is to fuse DL representations with expert knowledge. One way to address this challenge, as well as the performance-explainability trade-off is by leveraging the best of both streams without obviating domain expert knowledge. In this paper, we tackle such problem by considering the symbolic knowledge is expressed in form of a domain expert knowledge graph. We present the eXplainable Neural-symbolic learning (X-NeSyL) methodology, designed to learn both symbolic and deep representations, together with an explainability metric to assess the level of alignment of machine and human expert explanations. The ultimate objective is to fuse DL representations with expert domain knowledge during the learning process so it serves as a sound basis for explainability. In particular, X-NeSyL methodology involves the concrete use of two notions of explanation, both at inference and training time respectively: 1) EXPLANet: Expert-aligned eXplainable Part-based cLAssifier NETwork Architecture, a compositional convolutional neural network that makes use of symbolic representations, and 2) SHAP-Backprop, an explainable AI-informed training procedure that corrects and guides the DL process to align with such symbolic representations in form of knowledge graphs. We showcase X-NeSyL methodology using MonuMAI dataset for monument facade image classification, and demonstrate that our approach improves not only explainability of DL models but also performance.

show abstract

“…With the increasing complexity of modern Neural Networks [15], it is a difficult task to explain what particular features do influence the prediction. Therefore, DNNs have often been considered as 'black-box' [16], [17]. However, especially in security-critical applications (such as autonomous driving or personalized medicine), transparency of the decision-making model is mandatory and therefore the network's inability to explain its predictions restricts the applicability of ML systems.…”

Section: Introductionmentioning

confidence: 99%

Explaining Bayesian Neural Networks

Bykov,

Höhne,

Creosteanu

et al. 2021

Preprint

View full text Add to dashboard Cite

To make advanced learning machines such as Deep Neural Networks (DNNs) more transparent in decision making, explainable AI (XAI) aims to provide interpretations of DNNs' predictions. These interpretations are usually given in the form of heatmaps, each one illustrating relevant patterns regarding the prediction for a given instance. Bayesian approaches such as Bayesian Neural Networks (BNNs) so far have a limited form of transparency (model transparency) already built-in through their prior weight distribution, but notably, they lack explanations of their predictions for given instances. In this work, we bring together these two perspectives of transparency into a holistic explanation framework for explaining BNNs. Within the Bayesian framework, the network weights follow a probability distribution. Hence, the standard (deterministic) prediction strategy of DNNs extends in BNNs to a predictive distribution, and thus the standard explanation extends to an explanation distribution. Exploiting this view, we uncover that BNNs implicitly employ multiple heterogeneous prediction strategies. While some of these are inherited from standard DNNs, others are revealed to us by considering the inherent uncertainty in BNNs. Our quantitative and qualitative experiments on toy/benchmark data and realworld data from pathology show that the proposed approach of explaining BNNs can lead to more effective and insightful explanations.

show abstract

Analysis of Explainers of Black Box Deep Neural Networks for Computer Vision: A Survey

Cited by 28 publications

References 62 publications

CDLNet: Noise-Adaptive Convolutional Dictionary Learning Network for Blind Denoising and Demosaicing

CDLNet: Noise-Adaptive Convolutional Dictionary Learning Network for Blind Denoising and Demosaicing

EXplainable Neural-Symbolic Learning (X-NeSyL) methodology to fuse deep learning representations with expert knowledge graphs: the MonuMAI cultural heritage use case

Explaining Bayesian Neural Networks

Contact Info

Product

Resources

About