<i>DeepVID</i>: Deep Visual Interpretation and Diagnosis for Image Classifiers via Knowledge Distillation

Wang, Junpeng; Gou, Liang; Zhang, Wei; Yang, Hao; Shen, Han‐Wei

doi:10.1109/tvcg.2019.2903943

Cited by 111 publications

(62 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is very hard to understand what happens in the hidden layers and why a trained NN gives a positive diagnosis for a given input sample. This ''black-box'' aspect is very restrictive in many application fields, where the interpretation of a decision can lead to serious legal consequences especially in safetycritical applications [14] (e.g., medical diagnosis [15], [16], autonomous driving, electric power generation. .…”

Section: Introductionmentioning

confidence: 99%

“…Feature maps of the model are then obtained. The t-SNE method has been used and reported in many research publications such as [8], [14], [21]- [25]. The main constraint of this method is the lack of repeatability due to the minimization of the Kullback-Leibler divergence between the input space distribution and the embedding space distribution [22].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Deep Convolutional Variational Autoencoder as a 2D-Visualization Tool for Partial Discharge Source Classification in Hydrogenerators

et al. 2020

View full text Add to dashboard Cite

Hydrogenerators are strategic assets for power utilities. Their reliability and availability can lead to significant benefits. For decades, monitoring and diagnosis of hydrogenerators have been at the core of maintenance strategies. A significant part of generator diagnosis relies on Partial Discharge (PD) measurements, because the main cause of hydrogenerator breakdown comes from failure of its high voltage stator, which is a major component of hydrogenerators. A study of all stator failure mechanisms reveals that more than 85 % of them involve the presence of PD activity. PD signal can be detected from the lead of the hydrogenerator while it is running, thus allowing for on-line diagnosis. Hydro-Québec has been collecting more than 33 000 unlabeled PD measurement files over the last decades. Up to now, this diagnostic technique has been quantified based on global PD amplitudes and integrated PD energy irrespective of the source of the PD signal. Several PD sources exist and they all have different relative risk, but in order to recognize the nature of the PD, or its source, the judgement of experts is required. In this paper, we propose a new method based on visual data analysis to build a PD source classifier with a minimum of labeled data. A convolutional variational autoencoder has been used to help experts to visually select the best training data set in order to improve the performances of the PD source classifier.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Deep Convolutional Variational Autoencoder as a 2D-Visualization Tool for Partial Discharge Source Classification in Hydrogenerators

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Hohman et al [29] presented a comprehensive survey to summarize the state-of-the-art visual analysis methods for explainable deep learning. Existing methods can be categorized into three classes: network-centric [30], [31], [32], instance-centric [20], [33], [34], [35], and hybrid [36], [37]. Network-centric methods.…”

Section: Related Workmentioning

confidence: 99%

Analyzing the Noise Robustness of Deep Neural Networks

Cao

Liu

et al. 2021

IEEE Trans. Visual. Comput. Graphics

View full text Add to dashboard Cite

A B C D E (a) Input images (b) Datapath visualization at the network-and layer-levels (c) Neuron visualization Figure 1: Explanation why an adversarial panda image is not classified as a panda. The root cause is identified as the neurons in the feature map F A failing to detect the outline of the panda's ear (E) in the adversarial example, which further leads to the failure of detecting the panda's ear (B) in F C1 . Abstract-Adversarial examples, generated by adding small but intentionally imperceptible perturbations to normal examples, can mislead deep neural networks (DNNs) to make incorrect predictions. Although much work has been done on both adversarial attack and defense, a fine-grained understanding of adversarial examples is still lacking. To address this issue, we present a visual analysis method to explain why adversarial examples are misclassified. The key is to compare and analyze the datapaths of both the adversarial and normal examples. A datapath is a group of critical neurons along with their connections. We formulate the datapath extraction as a subset selection problem and solve it by constructing and training a neural network. A multi-level visualization consisting of a network-level visualization of data flows, a layer-level visualization of feature maps, and a neuron-level visualization of learned features, has been designed to help investigate how datapaths of adversarial and normal examples diverge and merge in the prediction process. A quantitative evaluation and a case study were conducted to demonstrate the promise of our method to explain the misclassification of adversarial examples.

show abstract

“…Liu et al [28] surveyed the recent progress on visualizations developed to understand, diagnose, and refine ML models. For example, Wang et al [50] developed an interpretation approach to review the inner mechanisms of complicated deep neural networks. Manifold [51] is a framework for visually interpreting, debugging, and comparing ML models.…”

Section: Visualizations For Xaimentioning

confidence: 99%

A visual analytics system for multi-model comparison on clinical data predictions

Fujiwara

Choi

et al. 2020

Visual Informatics

View full text Add to dashboard Cite

There is a growing trend of applying machine learning methods to medical datasets in order to predict patients' future status. Although some of these methods achieve high performance, challenges still exist in comparing and evaluating different models through their interpretable information. Such analytics can help clinicians improve evidence-based medical decision making. In this work, we develop a visual analytics system that compares multiple models' prediction criteria and evaluates their consistency. With our system, users can generate knowledge on different models' inner criteria and how confidently we can rely on each model's prediction for a certain patient. Through a case study of a publicly available clinical dataset, we demonstrate the effectiveness of our visual analytics system to assist clinicians and researchers in comparing and quantitatively evaluating different machine learning methods.

show abstract

DeepVID: Deep Visual Interpretation and Diagnosis for Image Classifiers via Knowledge Distillation

Cited by 111 publications

References 22 publications

Deep Convolutional Variational Autoencoder as a 2D-Visualization Tool for Partial Discharge Source Classification in Hydrogenerators

Deep Convolutional Variational Autoencoder as a 2D-Visualization Tool for Partial Discharge Source Classification in Hydrogenerators

Analyzing the Noise Robustness of Deep Neural Networks

A visual analytics system for multi-model comparison on clinical data predictions

Contact Info

Product

Resources

About