Search Results for author:
Found 37 papers, 12 papers with code
Layer-wise Feedback Propagation
In this paper, we present Layer-wise Feedback Propagation (LFP), a novel training approach for neural-network-like predictors that utilizes explainability, specifically Layer-wise Relevance Propagation(LRP), to assign rewards to individual connections based on their respective contributions to solving a given task.
Optimizing Explanations by Network Canonization and Hyperparameter Search
We further suggest a XAI evaluation framework with which we quantify and compare the effect sof model canonization for various XAI methods in image classification tasks on the Pascal-VOC and ILSVRC2017 datasets, as well as for Visual Question Answering using CLEVR-XAI.
Explainable Artificial Intelligence (XAI)
Image Classification
+2
Shortcomings of Top-Down Randomization-Based Sanity Checks for Evaluations of Deep Neural Network Explanations
To address shortcomings of this test, we start by observing an experimental gap in the ranking of explanation methods between randomization-based sanity checks [1] and model output faithfulness measures (e. g. [25]).
Discovering Transferable Forensic Features for CNN-generated Images Detection
Visual counterfeits are increasingly causing an existential conundrum in mainstream media with rapid evolution in neural image synthesis methods.
Beyond Explaining: Opportunities and Challenges of XAI-Based Model Improvement
We conclude that while model improvement based on XAI can have significant beneficial effects even on complex and not easily quantifyable model properties, these methods need to be applied carefully, since their success can vary depending on a multitude of factors, such as the model and dataset used, or the employed explanation method.
Explainable artificial intelligence
Explainable Artificial Intelligence (XAI)
Towards A Conceptually Simple Defensive Approach for Few-shot classifiers Against Adversarial Support Samples
In this work, we aim to close this gap by studying a conceptually simple approach to defend few-shot classifiers against adversarial attacks.
On the Robustness of Pretraining and Self-Supervision for a Deep Learning-based Analysis of Diabetic Retinopathy
Our results indicate that models initialized from ImageNet pretraining report a significant increase in performance, generalization and robustness to image distortions.
Detection of Adversarial Supports in Few-shot Classifiers Using Self-Similarity and Filtering
In this work, we propose a detection strategy to identify adversarial support sets, aimed at destroying the understanding of a few-shot classifier for a certain class.
Toward Scalable and Unified Example-based Explanation and Outlier Detection
We compare performances in terms of the classification, explanation quality, and outlier detection of our proposed network with other baselines.
Split and Expand: An inference-time improvement for Weakly Supervised Cell Instance Segmentation
In this work, we propose a two-step post-processing procedure, Split and Expand, that directly improves the conversion of segmentation maps to instances.
Explanation-Guided Training for Cross-Domain Few-Shot Classification
It leverages on the explanation scores, obtained from existing explanation methods when applied to the predictions of FSC models, computed for intermediate feature maps of the models.
Ranked #8 on
Cross-Domain Few-Shot
on ISIC2018
Understanding Integrated Gradients with SmoothTaylor for Deep Neural Network Attribution
From our experiments, we find that the SmoothTaylor approach together with adaptive noising is able to generate better quality saliency maps with lesser noise and higher sensitivity to the relevant points in the input space as compared to Integrated Gradients.
SideInfNet: A Deep Neural Network for Semi-Automatic Semantic Segmentation with Side Information
Fully-automatic execution is the ultimate goal for many Computer Vision applications.
Simple and Effective Prevention of Mode Collapse in Deep One-Class Classification
However, deep SVDD suffers from hypersphere collapse -- also known as mode collapse, if the architecture of the model does not comply with certain architectural constraints, e. g. the removal of bias terms.
Explain and Improve: LRP-Inference Fine-Tuning for Image Captioning Models
We develop variants of layer-wise relevance propagation (LRP) and gradient-based explanation methods, tailored to image captioning models with attention mechanisms.
Pruning by Explaining: A Novel Criterion for Deep Neural Network Pruning
The success of convolutional neural networks (CNNs) in various applications is accompanied by a significant increase in computation and parameter storage costs.
Explainable Artificial Intelligence (XAI)
Model Compression
+2
Exploring the Back Alleys: Analysing The Robustness of Alternative Neural Network Architectures against Adversarial Attacks
We investigate to what extent alternative variants of Artificial Neural Networks (ANNs) are susceptible to adversarial attacks.
Towards Best Practice in Explaining Neural Network Decisions with LRP
In this paper, we focus on a popular and widely used method of XAI, the Layer-wise Relevance Propagation (LRP).
Ranked #1 on
Object Detection
on SIXray
Explainable artificial intelligence
Explainable Artificial Intelligence (XAI)
+3
Resolving challenges in deep learning-based analyses of histopathological images using explanation methods
Deep learning has recently gained popularity in digital pathology due to its high prediction quality.
Deep Semi-Supervised Anomaly Detection
Deep approaches to anomaly detection have recently shown promising results over shallow methods on large and complex datasets.
Adversarial Attacks on Remote User Authentication Using Behavioural Mouse Dynamics
In an attempt to address this gap, we built a set of attacks, which are applications of several generative approaches, to construct adversarial mouse trajectories that bypass authentication models.
Unmasking Clever Hans Predictors and Assessing What Machines Really Learn
Current learning machines have successfully solved hard application problems, reaching high accuracy and displaying seemingly "intelligent" behavior.
SmartOTPs: An Air-Gapped 2-Factor Authentication for Smart-Contract Wallets (Extended Version)
The proposed framework consists of four components (i. e., an authenticator, a client, a hardware wallet, and a smart contract), and it provides 2-factor authentication (2FA) performed in two stages of interaction with the blockchain.
Cryptography and Security
Urban Zoning Using Higher-Order Markov Random Fields on Multi-View Imagery Data
Urban zoning enables various applications in land use analysis and urban planning.
Deep One-Class Classification
Despite the great advances made by deep learning in many machine learning problems, there is a relative dearth of deep learning approaches for anomaly detection.
Ranked #32 on
Anomaly Detection
on One-class CIFAR-10
Towards computational fluorescence microscopy: Machine learning-based integrated prediction of morphological and molecular tumor profiles
Recent advances in cancer research largely rely on new developments in microscopic or molecular profiling techniques offering high level of detail with respect to either spatial or molecular features, but usually not both.
Understanding and Comparing Deep Neural Networks for Age and Gender Classification
Recently, deep neural networks have demonstrated excellent performances in recognizing the age and gender on human face images.
Interpreting the Predictions of Complex ML Models by Layer-wise Relevance Propagation
Complex nonlinear models such as deep neural network (DNNs) have become an important tool for image classification, speech recognition, natural language processing, and many other fields of application.
Object Boundary Detection and Classification with Image-level Labels
We propose a novel strategy for solving this task, when pixel-level annotations are not available, performing it in an almost zero-shot manner by relying on conventional whole image neural net classifiers that were trained using large bounding boxes.
Layer-wise Relevance Propagation for Neural Networks with Local Renormalization Layers
Layer-wise relevance propagation is a framework which allows to decompose the prediction of a deep neural network computed over a sample, e. g. an image, down to relevance scores for the single input dimensions of the sample such as subpixels of an image.
Controlling Explanatory Heatmap Resolution and Semantics via Decomposition Depth
We present an application of the Layer-wise Relevance Propagation (LRP) algorithm to state of the art deep convolutional neural networks and Fisher Vector classifiers to compare the image perception and prediction strategies of both classifiers with the use of visualized heatmaps.
Explaining NonLinear Classification Decisions with Deep Taylor Decomposition
Although our focus is on image classification, the method is applicable to a broad set of input data, learning tasks and network architectures.
Analyzing Classifiers: Fisher Vectors and Deep Neural Networks
Fisher Vector classifiers and Deep Neural Networks (DNNs) are popular and successful algorithms for solving image classification problems.
Evaluating the visualization of what a Deep Neural Network has learned
Our main result is that the recently proposed Layer-wise Relevance Propagation (LRP) algorithm qualitatively and quantitatively provides a better explanation of what made a DNN arrive at a particular classification decision than the sensitivity-based approach or the deconvolution method.
Localized Multiple Kernel Learning---A Convex Approach
We propose a localized approach to multiple kernel learning that can be formulated as a convex optimization problem over a given cluster structure.
Multi-class SVMs: From Tighter Data-Dependent Generalization Bounds to Novel Algorithms
This paper studies the generalization performance of multi-class classification algorithms, for which we obtain, for the first time, a data-dependent generalization error bound with a logarithmic dependence on the class size, substantially improving the state-of-the-art linear dependence in the existing data-dependent generalization analysis.
Multiple Kernel Learning for Brain-Computer Interfacing
Combining information from different sources is a common way to improve classification accuracy in Brain-Computer Interfacing (BCI).
