arXiv subjects

Jindong Gu

Publications and source records attributed to Jindong Gu.

130 records · Page 8Linked to original sources

arXiv2019

Saliency Methods for Explaining Adversarial Attacks

The classification decisions of neural networks can be misled by small imperceptible perturbations. This work aims to explain the misled classifications using saliency methods. The idea behind saliency methods is to explain the classification decisions of neural networks by creating so-called saliency maps. Unfortunately, a number of recent publications have shown that many of the proposed saliency methods do not provide insightful explanations. A prominent example is Guided Backpropagation (GuidedBP), which simply performs (partial) image recovery. However, our numerical analysis shows the saliency maps created by GuidedBP do indeed contain class-discriminative information. We propose a simple and efficient way to enhance the saliency maps. The proposed enhanced GuidedBP shows the state-of-the-art performance to explain adversary classifications.

cs.CV↗

arXiv2019

Semantics for Global and Local Interpretation of Deep Neural Networks

Deep neural networks (DNNs) with high expressiveness have achieved state-of-the-art performance in many tasks. However, their distributed feature representations are difficult to interpret semantically. In this work, human-interpretable semantic concepts are associated with vectors in feature space. The association process is mathematically formulated as an optimization problem. The semantic vectors obtained from the optimal solution are applied to interpret deep neural networks globally and locally. The global interpretations are useful to understand the knowledge learned by DNNs. The interpretation of local behaviors can help to understand individual decisions made by DNNs better. The empirical experiments demonstrate how to use identified semantics to interpret the existing DNNs.

cs.CV↗

arXiv2019

Understanding Individual Decisions of CNNs via Contrastive Backpropagation

A number of backpropagation-based approaches such as DeConvNets, vanilla Gradient Visualization and Guided Backpropagation have been proposed to better understand individual decisions of deep convolutional neural networks. The saliency maps produced by them are proven to be non-discriminative. Recently, the Layer-wise Relevance Propagation (LRP) approach was proposed to explain the classification decisions of rectifier neural networks. In this work, we evaluate the discriminativeness of the generated explanations and analyze the theoretical foundation of LRP, i.e. Deep Taylor Decomposition. The experiments and analysis conclude that the explanations generated by LRP are not class-discriminative. Based on LRP, we propose Contrastive Layer-wise Relevance Propagation (CLRP), which is capable of producing instance-specific, class-discriminative, pixel-wise explanations. In the experiments, we use the CLRP to explain the decisions and understand the difference between neurons in individual classification decisions. We also evaluate the explanations quantitatively with a Pointing Game and an ablation study. Both qualitative and quantitative evaluations show that the CLRP generates better explanations than the LRP. The code is available.

cs.CV↗

arXiv2019

Understanding Bias in Machine Learning

Bias is known to be an impediment to fair decisions in many domains such as human resources, the public sector, health care etc. Recently, hope has been expressed that the use of machine learning methods for taking such decisions would diminish or even resolve the problem. At the same time, machine learning experts warn that machine learning models can be biased as well. In this article, our goal is to explain the issue of bias in machine learning from a technical perspective and to illustrate the impact that biased data can have on a machine learning model. To reach such a goal, we develop interactive plots to visualizing the bias learned from synthetic data.

cs.LG↗