Integrated Grad-Cam: Sensitivity-Aware Visual Explanation of Deep Convolutional Networks Via Integrated Gradient-Based Scoring

Sattarzadeh, Sam; Sudhakar, M.; Plataniotis, Konstantinos N.; Jang, Ji Hye; Jeong, Yeonjeong; Kim, H

doi:10.1109/icassp39728.2021.9415064

Cited by 24 publications

(16 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Optimization related methods, including Feedback [38] and FGVis [29], add the complex switch structure into the network and iteratively optimize the object function to achieve the saliency maps. Some integration related methods, such as SmoothGrad [39], IntegratedGrad [40], and Integrated Grad-CAM [41], can be regarded as packages over single propagation methods. We can also take advantages of these packaging tools to increment our method effects.…”

Section: B Iteration Based Methodsmentioning

confidence: 99%

TSGB: Target-Selective Gradient Backprop for Probing CNN Visual Saliency

Cheng,

Fang,

Liang

et al. 2021

Preprint

View full text Add to dashboard Cite

The explanation for deep neural networks has drawn extensive attention in the deep learning community over the past few years. In this work, we study the visual saliency, a.k.a. visual explanation, to interpret convolutional neural networks. Compared to iteration based saliency methods, single backward pass based saliency methods benefit from faster speed and are widely used in downstream visual tasks. Thus our work focuses on single backward pass approaches. However, existing methods in this category struggle to successfully produce finegrained saliency maps concentrating on specific target classes. That said, producing faithful saliency maps satisfying both targetselectiveness and fine-grainedness using a single backward pass is a challenging problem in the field. To mitigate this problem, we revisit the gradient flow inside the network, and find that the entangled semantics and original weights may disturb the propagation of target-relevant saliency. Inspired by those observations, we propose a novel visual saliency framework, termed Target-Selective Gradient (TSG) backprop, which leverages rectification operations to effectively emphasize target classes and further efficiently propagate the saliency to the input space, thereby generating target-selective and fine-grained saliency maps. The proposed TSG consists of two components, namely, TSG-Conv and TSG-FC, which rectify the gradients for convolutional layers and fully-connected layers, respectively. Thorough qualitative and quantitative experiments on ImageNet and Pascal VOC show that the proposed framework achieves more accurate and reliable results than other competitive methods.

show abstract

Section: B Iteration Based Methodsmentioning

confidence: 99%

TSGB: Target-Selective Gradient Backprop for Probing CNN Visual Saliency

Cheng,

Fang,

Liang

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…In this section we explain how these methods work. We also look at the Integrated Grad-CAM technique introduced in [Sattarzadeh 2021], and show how it differs from ours.…”

Section: Previous Workmentioning

confidence: 99%

“…In our study we look at three gradient based meth-ods: Gradient Guided Class Activation Map (Grad-CAM, a layer attribution method) [Selvaraju 2017], Integrated Gradients (a primary attribution method) [Sundararajan 2017], and Integrated Grad-CAM (layer attribution) [Sattarzadeh 2021]. We examine their advantages and limitations, and propose a modification of Grad-CAM in which gradients are replaced with integration of gradients computed at any layer rather than the input layer.…”

Section: Introductionmentioning

confidence: 99%

Visual Explanations from Deep Networks via Riemann-Stieltjes Integrated Gradient-based Localization

Lucas¹,

Miguel²,

Furst³

et al. 2022

Preprint

View full text Add to dashboard Cite

Neural networks are becoming increasingly better at tasks that involve classifying and recognizing images. At the same time techniques intended to explain the network output have been proposed. One such technique is the Gradient-based Class Activation Map (Grad-CAM), which is able to locate features of an input image at various levels of a convolutional neural network (CNN), but is sensitive to the vanishing gradients problem. There are techniques such as Integrated Gradients (IG), that are not affected by that problem, but its use is limited to the input layer of a network. Here we introduce a new technique to produce visual explanations for the predictions of a CNN. Like Grad-CAM, our method can be applied to any layer of the network, and like Integrated Gradients it is not affected by the problem of vanishing gradients. For efficiency, gradient integration is performed numerically at the layer level using a Riemann-Stieltjes sum approximation. Compared to Grad-CAM, heatmaps produced by our algorithm are better focused in the areas of interest, and their numerical computation is more stable. Our code is available at https://github.com/mlerma54/RSIGradCAM.

show abstract

“…The weights indicate the importance of the target category c for the feature map. Since the focus is only on the influence of positive values in the feature map on the final classification result, it is necessary to remove the influence of negative values using a further ReLU function on the resulting feature map weighted by the weights [ 15 ]. The output map of Grad-CAM can be obtained after this weighting is done as follows: …”

Section: Introductionmentioning

confidence: 99%

Side channel analysis based on feature fusion network

Wang²,

Tang³

et al. 2022

PLoS ONE

View full text Add to dashboard Cite

Various physical information can be leaked while the encryption algorithm is running in the device. Side-channel analysis exploits these leakages to recover keys. Due to the sensitivity of deep learning to the data features, the efficiency and accuracy of side channel analysis are effectively improved with the application of deep learning algorithms. However, a considerable part of existing reserches are based on traditional neural networks. The effectiveness of key recovery is improved by increasing the size of the network. However, the computational complexity of the algorithm increases accordingly. Problems such as overfitting, low training efficiency, and low feature extraction ability also occur. In this paper, we construct an improved lightweight convolutional neural network based on the feature fusion network. The new network and the traditional neural networks are respectively applied to the side-channel analysis for comparative experiments. The results show that the new network has faster convergence, better robustness and higher accuracy. No overfitting has occurred. A heatmap visualization method was introduced for analysis. The new network has higher heat value and more concentration in the key interval. Side-channel analysis based on feature fusion network has better performance, compared with the ones based on traditional neural networks.

show abstract

Integrated Grad-Cam: Sensitivity-Aware Visual Explanation of Deep Convolutional Networks Via Integrated Gradient-Based Scoring

Cited by 24 publications

References 10 publications

TSGB: Target-Selective Gradient Backprop for Probing CNN Visual Saliency

TSGB: Target-Selective Gradient Backprop for Probing CNN Visual Saliency

Visual Explanations from Deep Networks via Riemann-Stieltjes Integrated Gradient-based Localization

Side channel analysis based on feature fusion network

Contact Info

Product

Resources

About