Information-Theoretic Visual Explanation for Black-Box Classifiers

• URL: http://arxiv.org/abs/2009.11150v2
• Date: Fri, 16 Jul 2021 07:40:24 GMT
• Title: Information-Theoretic Visual Explanation for Black-Box Classifiers
• Authors: Jihun Yi, Eunji Kim, Siwon Kim, Sungroh Yoon
• Abstract summary: In this work, we attempt to explain the prediction of any black-box classifier from an information-theoretic perspective. We obtain two attribution maps--an information gain (IG) map and a point-wise mutual information (PMI) map. Compared to existing methods, our method improves the correctness of the attribution maps in terms of a quantitative metric.
• Score: 30.62290460123988
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: In this work, we attempt to explain the prediction of any black-box classifier from an information-theoretic perspective. For each input feature, we compare the classifier outputs with and without that feature using two information-theoretic metrics. Accordingly, we obtain two attribution maps--an information gain (IG) map and a point-wise mutual information (PMI) map. IG map provides a class-independent answer to "How informative is each pixel?", and PMI map offers a class-specific explanation of "How much does each pixel support a specific class?" Compared to existing methods, our method improves the correctness of the attribution maps in terms of a quantitative metric. We also provide a detailed analysis of an ImageNet classifier using the proposed method, and the code is available online.

Related papers

• Now you see me! A framework for obtaining class-relevant saliency maps [38.663697418404546]
  Saliency maps have been developed to gain understanding into which input features neural networks use for a specific prediction. Although widely employed, these methods often result in overly general saliency maps that fail to identify the specific information that triggered the classification. We suggest a framework that allows to incorporate attributions across classes to arrive at saliency maps that actually capture the class-relevant information.
  arXiv  Detail & Related papers  (2025-03-10T13:59:57Z)
• Taming CLIP for Fine-grained and Structured Visual Understanding of Museum Exhibits [59.66134971408414]
  We aim to adapt CLIP for fine-grained and structured understanding of museum exhibits. Our dataset is the first of its kind in the public domain. The proposed method (MUZE) learns to map CLIP's image embeddings to the tabular structure by means of a proposed transformer-based parsing network (parseNet)
  arXiv  Detail & Related papers  (2024-09-03T08:13:06Z)
• Masked Image Modeling: A Survey [73.21154550957898]
  Masked image modeling emerged as a powerful self-supervised learning technique in computer vision. We construct a taxonomy and review the most prominent papers in recent years. We aggregate the performance results of various masked image modeling methods on the most popular datasets.
  arXiv  Detail & Related papers  (2024-08-13T07:27:02Z)
• Accurate Explanation Model for Image Classifiers using Class Association Embedding [5.378105759529487]
  We propose a generative explanation model that combines the advantages of global and local knowledge. Class association embedding (CAE) encodes each sample into a pair of separated class-associated and individual codes. Building-block coherency feature extraction algorithm is proposed that efficiently separates class-associated features from individual ones.
  arXiv  Detail & Related papers  (2024-06-12T07:41:00Z)
• Interpretable Network Visualizations: A Human-in-the-Loop Approach for Post-hoc Explainability of CNN-based Image Classification [5.087579454836169]
  State-of-the-art explainability methods generate saliency maps to show where a specific class is identified. We introduce a post-hoc method that explains the entire feature extraction process of a Convolutional Neural Network. We also show an approach to generate global explanations by aggregating labels across multiple images.
  arXiv  Detail & Related papers  (2024-05-06T09:21:35Z)
• DXAI: Explaining Classification by Image Decomposition [4.013156524547072]
  We propose a new way to visualize neural network classification through a decomposition-based explainable AI (DXAI) Instead of providing an explanation heatmap, our method yields a decomposition of the image into class-agnostic and class-distinct parts.
  arXiv  Detail & Related papers  (2023-12-30T20:52:20Z)
• Feature Activation Map: Visual Explanation of Deep Learning Models for Image Classification [17.373054348176932]
  In this work, a post-hoc interpretation tool named feature activation map (FAM) is proposed. FAM can interpret deep learning models without FC layers as a classifier. Experiments conducted on ten deep learning models for few-shot image classification, contrastive learning image classification and image retrieval tasks demonstrate the effectiveness of the proposed FAM algorithm.
  arXiv  Detail & Related papers  (2023-07-11T05:33:46Z)
• Text Descriptions are Compressive and Invariant Representations for Visual Learning [63.3464863723631]
  We show that an alternative approach, in line with humans' understanding of multiple visual features per class, can provide compelling performance in the robust few-shot learning setting. In particular, we introduce a novel method, textit SLR-AVD (Sparse Logistic Regression using Augmented Visual Descriptors). This method first automatically generates multiple visual descriptions of each class via a large language model (LLM), then uses a VLM to translate these descriptions to a set of visual feature embeddings of each image, and finally uses sparse logistic regression to select a relevant subset of these features to classify
  arXiv  Detail & Related papers  (2023-07-10T03:06:45Z)
• Measuring the Interpretability of Unsupervised Representations via Quantized Reverse Probing [97.70862116338554]
  We investigate the problem of measuring interpretability of self-supervised representations. We formulate the latter as estimating the mutual information between the representation and a space of manually labelled concepts. We use our method to evaluate a large number of self-supervised representations, ranking them by interpretability.
  arXiv  Detail & Related papers  (2022-09-07T16:18:50Z)
• Learning Implicit Feature Alignment Function for Semantic Segmentation [51.36809814890326]
  Implicit Feature Alignment function (IFA) is inspired by the rapidly expanding topic of implicit neural representations. We show that IFA implicitly aligns the feature maps at different levels and is capable of producing segmentation maps in arbitrary resolutions. Our method can be combined with improvement on various architectures, and it achieves state-of-the-art accuracy trade-off on common benchmarks.
  arXiv  Detail & Related papers  (2022-06-17T09:40:14Z)
• LEAD: Self-Supervised Landmark Estimation by Aligning Distributions of Feature Similarity [49.84167231111667]
  Existing works in self-supervised landmark detection are based on learning dense (pixel-level) feature representations from an image. We introduce an approach to enhance the learning of dense equivariant representations in a self-supervised fashion. We show that having such a prior in the feature extractor helps in landmark detection, even under drastically limited number of annotations.
  arXiv  Detail & Related papers  (2022-04-06T17:48:18Z)
• Isometric Propagation Network for Generalized Zero-shot Learning [72.02404519815663]
  A popular strategy is to learn a mapping between the semantic space of class attributes and the visual space of images based on the seen classes and their data. We propose Isometric propagation Network (IPN), which learns to strengthen the relation between classes within each space and align the class dependency in the two spaces. IPN achieves state-of-the-art performance on three popular Zero-shot learning benchmarks.
  arXiv  Detail & Related papers  (2021-02-03T12:45:38Z)
• One Explanation is Not Enough: Structured Attention Graphs for Image Classification [30.640374946654806]
  We introduce structured attention graphs (SAGs), which compactly represent sets of attention maps for an image. We propose an approach to compute SAGs and a visualization for SAGs so that deeper insight can be gained into a classifier's decisions.
  arXiv  Detail & Related papers  (2020-11-13T02:51:54Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.