PyTorchKR Papers

“Image Classification”

총 18,403편
태그·기간 등 상세 필터

카테고리 탐색

arXiv:2403.00890

Improving Android Malware Detection Through Data Augmentation Using Wasserstein Generative Adversarial Networks

Generative Adversarial Networks (GANs) have demonstrated their versatility across various applications, including data augmentation and malware detection. This research explores the effectiveness of utilizing GAN-generated data to train a model for the detection of Android malware. Given the considerable storage requirements of Android applications, the study proposes a method to synthetically represent data using GANs, thereby reducing storage demands. The proposed methodology involves creating image representations of features extracted from an existing dataset. A GAN model is then employed to generate a more extensive dataset consisting of realistic synthetic grayscale images. Subsequently, this synthetic dataset is utilized to train a Convolutional Neural Network (CNN) designed to identify previously unseen Android malware applications. The study includes a comparative analysis of the CNN's performance when trained on real images versus synthetic images generated by the GAN. Furthermore, the research explores variations in performance between the Wasserstein Generative Adversarial Network (WGAN) and the Deep Convolutional Generative Adversarial Network (DCGAN). The investigation extends to studying the impact of image size and malware obfuscation on the classification model's effectiveness. The data augmentation approach implemented in this study resulted in a notable performance enhancement of the classification model, ranging from 1.5% to 7%, depending on the dataset. The highest achieved F1 score reached 0.975. Keywords--Generative Adversarial Networks, Android Malware, Data Augmentation, Wasserstein Generative Adversarial Network

arXiv:2409.06290공식 코드 · Papers with Code

EntAugment: Entropy-Driven Adaptive Data Augmentation Framework for Image Classification

Data augmentation (DA) has been widely used to improve the generalization of deep neural networks. While existing DA methods have proven effective, they often rely on augmentation operations with random magnitudes to each sample. However, this approach can inadvertently introduce noise, induce distribution shifts, and increase the risk of overfitting. In this paper, we propose EntAugment, a tuning-free and adaptive DA framework. Unlike previous work, EntAugment dynamically assesses and adjusts the augmentation magnitudes for each sample during training, leveraging insights into both the inherent complexities of training samples and the evolving status of deep models. Specifically, in EntAugment, the magnitudes are determined by the information entropy derived from the probability distribution obtained by applying the softmax function to the model's output. In addition, to further enhance the efficacy of EntAugment, we introduce a novel entropy regularization term, EntLoss, which complements the EntAugment approach. Theoretical analysis further demonstrates that EntLoss, compared to traditional cross-entropy loss, achieves closer alignment between the model distributions and underlying dataset distributions. Moreover, EntAugment and EntLoss can be utilized separately or jointly. We conduct extensive experiments across multiple image classification tasks and network architectures with thorough comparisons of existing DA methods. Importantly, the proposed methods outperform others without introducing any auxiliary models or noticeable extra computational costs, highlighting both effectiveness and efficiency. Code is available at https://github.com/Jackbrocp/EntAugment.

arXiv:2607.12556v1cs.CV

CGRL: 전체 슬라이드 이미지 분류를 위한 개념 유도 가지치기 및 표현 학습

CGRL: Concept-Guided Pruning and Representation Learning for Whole-Slide Image Classification

약한 지도 학습 기반 전체 슬라이드 이미지(WSI) 분류는 밀집 영역 주석보다 슬라이드 수준 레이블을 얻기가 더 쉬워 계산 병리학(computational pathology)에서 널리 사용된다. 기존 다중 인스턴스 학습(MIL) 방법은 주로 시각적 단서를 이용해 대량의 패치 임베딩 bag을 집계하는데, 이 과정에서 정보가 적은 패치를 많이 남기고 인스턴스 특징과 클래스 수준 질병 의미론 간의 정렬(alignment)이 약해질 수 있다. 우리는 개념 유도 가지치기 및 표현 학습(CGRL)을 제안한다. 이는 질병 프롬프트에서 유도된 클래스 수준 개념 프로토타입을 MIL 파이프라인에 도입하는 단순한 프레임워크이다. 첫째, 개념 관련성 가지치기는 패치 인스턴스를 클래스 개념과의 유사도에 따라 순위를 매기고, 하류 MIL 집계를 위해 상위 K개의 개념 관련 패치를 유지한다. 둘째, 개념 유도 대조 표현 학습은 동일한 유사도 행렬로부터 클래스별 양성 및 음성 패치 집합을 구성하고, 대상 클래스, 대칭 보조, 교차 클래스 분리 목표를 최적화하여 투영된 개념 공간을 정규화한다. 우리는 여러 대표적인 MIL 방법을 사용해 TCGA-BRCA 및 TCGA-NSCLC에서 CGRL을 평가한다. 실험 결과, CGRL은 여러 모델-데이터셋 조합을 개선하며, 향상 정도는 하류 MIL 모델과 데이터셋에 따라 달라진다. 개념 관련성 가지치기를 통해 계산 비용을 줄이면서 정확도와 macro-F1에서 특히 뚜렷한 향상을 달성한다. 이 결과는 클래스 수준 의미 개념이 약한 지도 학습 계산 병리학에서 패치 선택과 표현 학습을 위한 효과적이고 실용적인 사전 지식(prior)을 제공함을 보여준다.

arXiv:2603.06578v3cs.CV

이미지 분류기로서의 멀티모달 대규모 언어 모델(MLLM)

Multimodal Large Language Models as Image Classifiers

멀티모달 대규모 언어 모델(MLLM)의 분류 성능은 평가 프로토콜과 정답(ground truth) 품질에 크게 의존한다. MLLM을 지도학습 모델 및 비전-언어 모델(VLM)과 비교한 연구들은 서로 상충하는 결론을 보고하는데, 우리는 이러한 상충이 성능을 과장하거나 과소평가하는 프로토콜에서 비롯됨을 보인다. 가장 흔히 사용되는 평가 프로토콜 전반에서, 우리는 핵심 문제들을 식별하고 수정한다: 제공된 클래스 목록 범위를 벗어나 폐기되는 모델 출력, 약한 객관식 오답지(distractor)로 인한 과장된 결과, 그리고 부실한 출력 매핑으로 인해서만 성능이 저하되는 개방형(open-world) 설정이 그것이다. 추가로 우리는 흔히 간과되는 설계 선택들 - 배치 크기, 이미지 순서, 텍스트 인코더 선택 - 의 영향을 정량화하여, 이들이 정확도에 상당한 영향을 미친다는 것을 보인다. ImageNet-1k의 625개 클래스에 대한 우리의 다중 레이블 재주석(multilabel reannotation)인 ReGT로 평가한 결과, MLLM은 수정된 레이블로부터 가장 큰 이득(최대 +10.8%)을 얻으며, 지도학습 모델과의 인지된 격차를 크게 좁히는 것으로 나타났다. 따라서 보고된 MLLM의 분류 성능 저하의 상당 부분은 모델 자체의 결함이 아니라 잡음이 많은 정답(ground truth)과 결함 있는 평가 프로토콜에서 비롯된 인공물(artifact)이다. 지도학습 신호에 덜 의존하는 모델일수록 주석(annotation) 품질에 더 민감한 것으로 나타났다. 마지막으로, 우리는 MLLM이 사람 주석자(annotator)를 보조할 수 있음을 보인다: 통제된 사례 연구에서, 주석자들은 어려운 사례의 약 50%에서 MLLM의 예측을 확인하거나 통합하였으며, 이는 대규모 데이터셋 구축(curation)에 대한 가능성을 보여준다. 본 연구는 Aiming for Perfect ImageNet-1k 프로젝트의 일부이며, 자세한 내용은 https://klarajanouskova.github.io/ImageNet/ 에서 확인할 수 있다.

arXiv:2003.11213

A New Multiple Max-pooling Integration Module and Cross Multiscale Deconvolution Network Based on Image Semantic Segmentation

To better retain the deep features of an image and solve the sparsity problem of the end-to-end segmentation model, we propose a new deep convolutional network model for medical image pixel segmentation, called MC-Net. The core of this network model consists of four parts, namely, an encoder network, a multiple max-pooling integration module, a cross multiscale deconvolution decoder network and a pixel-level classification layer. In the network structure of the encoder, we use multiscale convolution instead of the traditional single-channel convolution. The multiple max-pooling integration module first integrates the output features of each submodule of the encoder network and reduces the number of parameters by convolution using a kernel size of 1. At the same time, each max-pooling layer (the pooling size of each layer is different) is spliced after each convolution to achieve the translation invariance of the feature maps of each submodule. We use the output feature maps from the multiple max-pooling integration module as the input of the decoder network; the multiscale convolution of each submodule in the decoder network is cross-fused with the feature maps generated by the corresponding multiscale convolution in the encoder network. Using the above feature map processing methods solves the sparsity problem after the max-pooling layer-generating matrix and enhances the robustness of the classification. We compare our proposed model with the well-known Fully Convolutional Networks for Semantic Segmentation (FCNs), DecovNet, PSPNet, U-net, SgeNet and other state-of-the-art segmentation networks such as HyperDenseNet, MS-Dual, Espnetv2, Denseaspp using one binary Kaggle 2018 data science bowl dataset and two multiclass dataset and obtain encouraging experimental results.

arXiv:2406.14583

Modeling & Evaluating the Performance of Convolutional Neural Networks for Classifying Steel Surface Defects

Recently, outstanding identification rates in image classification tasks were achieved by convolutional neural networks (CNNs). to use such skills, selective CNNs trained on a dataset of well-known images of metal surface defects captured with an RGB camera. Defects must be detected early to take timely corrective action due to production concerns. For image classification up till now, a model-based method has been utilized, which indicated the predicted reflection characteristics of surface defects in comparison to flaw-free surfaces. The problem of detecting steel surface defects has grown in importance as a result of the vast range of steel applications in end-product sectors such as automobiles, households, construction, etc. The manual processes for detections are time-consuming, labor-intensive, and expensive. Different strategies have been used to automate manual processes, but CNN models have proven to be the most effective rather than image processing and machine learning techniques. By using different CNN models with fine-tuning, easily compare their performance and select the best-performing model for the same kinds of tasks. However, it is important that using different CNN models either from fine tuning can be computationally expensive and time-consuming. Therefore, our study helps the upcoming researchers to choose the CNN without considering the issues of model complexity, performance, and computational resources. In this article, the performance of various CNN models with transfer learning techniques are evaluated. These models were chosen based on their popularity and impact in the field of computer vision research, as well as their performance on benchmark datasets. According to the outcomes, DenseNet201 outperformed the other CNN models and had the greatest detection rate on the NEU dataset, falling in at 98.37 percent.

arXiv:2005.14117

Multimodal Feature Fusion and Knowledge-Driven Learning via Experts Consult for Thyroid Nodule Classification

Computer-aided diagnosis (CAD) is becoming a prominent approach to assist clinicians spanning across multiple fields. These automated systems take advantage of various computer vision (CV) procedures, as well as artificial intelligence (AI) techniques, to formulate a diagnosis of a given image, e.g., computed tomography and ultrasound. Advances in both areas (CV and AI) are enabling ever increasing performances of CAD systems, which can ultimately avoid performing invasive procedures such as fine-needle aspiration. In this study, a novel end-to-end knowledge-driven classification framework is presented. The system focuses on multimodal data generated by thyroid ultrasonography, and acts as a CAD system by providing a thyroid nodule classification into the benign and malignant categories. Specifically, the proposed system leverages cues provided by an ensemble of experts to guide the learning phase of a densely connected convolutional network (DenseNet). The ensemble is composed by various networks pretrained on ImageNet, including AlexNet, ResNet, VGG, and others. The previously computed multimodal feature parameters are used to create ultrasonography domain experts via transfer learning, decreasing, moreover, the number of samples required for training. To validate the proposed method, extensive experiments were performed, providing detailed performances for both the experts ensemble and the knowledge-driven DenseNet. As demonstrated by the results, the proposed system achieves relevant performances in terms of qualitative metrics for the thyroid nodule classification task, thus resulting in a great asset when formulating a diagnosis.

arXiv:2106.02566공식 코드 · Papers with Code

BR-NPA: A Non-Parametric High-Resolution Attention Model to improve the Interpretability of Attention

The prevalence of employing attention mechanisms has brought along concerns on the interpretability of attention distributions. Although it provides insights about how a model is operating, utilizing attention as the explanation of model predictions is still highly dubious. The community is still seeking more interpretable strategies for better identifying local active regions that contribute the most to the final decision. To improve the interpretability of existing attention models, we propose a novel Bilinear Representative Non-Parametric Attention (BR-NPA) strategy that captures the task-relevant human-interpretable information. The target model is first distilled to have higher-resolution intermediate feature maps. From which, representative features are then grouped based on local pairwise feature similarity, to produce finer-grained, more precise attention maps highlighting task-relevant parts of the input. The obtained attention maps are ranked according to the activity level of the compound feature, which provides information regarding the important level of the highlighted regions. The proposed model can be easily adapted in a wide variety of modern deep models, where classification is involved. Extensive quantitative and qualitative experiments showcase more comprehensive and accurate visual explanations compared to state-of-the-art attention models and visualizations methods across multiple tasks including fine-grained image classification, few-shot classification, and person re-identification, without compromising the classification accuracy. The proposed visualization model sheds imperative light on how neural networks `pay their attention' differently in different tasks.

arXiv:2203.16000공식 코드 · Papers with Code

StyleFool: Fooling Video Classification Systems via Style Transfer

Video classification systems are vulnerable to adversarial attacks, which can create severe security problems in video verification. Current black-box attacks need a large number of queries to succeed, resulting in high computational overhead in the process of attack. On the other hand, attacks with restricted perturbations are ineffective against defenses such as denoising or adversarial training. In this paper, we focus on unrestricted perturbations and propose StyleFool, a black-box video adversarial attack via style transfer to fool the video classification system. StyleFool first utilizes color theme proximity to select the best style image, which helps avoid unnatural details in the stylized videos. Meanwhile, the target class confidence is additionally considered in targeted attacks to influence the output distribution of the classifier by moving the stylized video closer to or even across the decision boundary. A gradient-free method is then employed to further optimize the adversarial perturbations. We carry out extensive experiments to evaluate StyleFool on two standard datasets, UCF-101 and HMDB-51. The experimental results demonstrate that StyleFool outperforms the state-of-the-art adversarial attacks in terms of both the number of queries and the robustness against existing defenses. Moreover, 50% of the stylized videos in untargeted attacks do not need any query since they can already fool the video classification model. Furthermore, we evaluate the indistinguishability through a user study to show that the adversarial samples of StyleFool look imperceptible to human eyes, despite unrestricted perturbations.

arXiv:2307.01951공식 코드 · Papers with Code

A Neural Collapse Perspective on Feature Evolution in Graph Neural Networks

Graph neural networks (GNNs) have become increasingly popular for classification tasks on graph-structured data. Yet, the interplay between graph topology and feature evolution in GNNs is not well understood. In this paper, we focus on node-wise classification, illustrated with community detection on stochastic block model graphs, and explore the feature evolution through the lens of the "Neural Collapse" (NC) phenomenon. When training instance-wise deep classifiers (e.g. for image classification) beyond the zero training error point, NC demonstrates a reduction in the deepest features' within-class variability and an increased alignment of their class means to certain symmetric structures. We start with an empirical study that shows that a decrease in within-class variability is also prevalent in the node-wise classification setting, however, not to the extent observed in the instance-wise case. Then, we theoretically study this distinction. Specifically, we show that even an "optimistic" mathematical model requires that the graphs obey a strict structural condition in order to possess a minimizer with exact collapse. Interestingly, this condition is viable also for heterophilic graphs and relates to recent empirical studies on settings with improved GNNs' generalization. Furthermore, by studying the gradient dynamics of the theoretical model, we provide reasoning for the partial collapse observed empirically. Finally, we present a study on the evolution of within- and between-class feature variability across layers of a well-trained GNN and contrast the behavior with spectral methods.

arXiv:1910.05611

Cross-Domain Image Classification through Neural-Style Transfer Data Augmentation

In particular, the lack of sufficient amounts of domain-specific data can reduce the accuracy of a classifier. In this paper, we explore the effects of style transfer-based data transformation on the accuracy of a convolutional neural network classifiers in the context of automobile detection under adverse winter weather conditions. The detection of automobiles under highly adverse weather conditions is a difficult task as such conditions present large amounts of noise in each image. The InceptionV2 architecture is trained on a composite dataset, consisting of either normal car image dataset , a mixture of normal and style transferred car images, or a mixture of normal car images and those taken at blizzard conditions, at a ratio of 80:20. All three classifiers are then tested on a dataset of car images taken at blizzard conditions and on vehicle-free snow landscape images. We evaluate and contrast the effectiveness of each classifier upon each dataset, and discuss the strengths and weaknesses of style-transfer based approaches to data augmentation.

arXiv:2006.07742

Split-Merge Pooling

There are a variety of approaches to obtain a vast receptive field with convolutional neural networks (CNNs), such as pooling or striding convolutions. Most of these approaches were initially designed for image classification and later adapted to dense prediction tasks, such as semantic segmentation. However, the major drawback of this adaptation is the loss of spatial information. Even the popular dilated convolution approach, which in theory is able to operate with full spatial resolution, needs to subsample features for large image sizes in order to make the training and inference tractable. In this work, we introduce Split-Merge pooling to fully preserve the spatial information without any subsampling. By applying Split-Merge pooling to deep networks, we achieve, at the same time, a very large receptive field. We evaluate our approach for dense semantic segmentation of large image sizes taken from the Cityscapes and GTA-5 datasets. We demonstrate that by replacing max-pooling and striding convolutions with our split-merge pooling, we are able to improve the accuracy of different variations of ResNet significantly.

arXiv:2007.15161공식 코드 · Papers with Code

Rethinking Recurrent Neural Networks and Other Improvements for Image Classification

Over the long history of machine learning, which dates back several decades, recurrent neural networks (RNNs) have been used mainly for sequential data and time series and generally with 1D information. Even in some rare studies on 2D images, these networks are used merely to learn and generate data sequentially rather than for image recognition tasks. In this study, we propose integrating an RNN as an additional layer when designing image recognition models. We also develop end-to-end multimodel ensembles that produce expert predictions using several models. In addition, we extend the training strategy so that our model performs comparably to leading models and can even match the state-of-the-art models on several challenging datasets (e.g., SVHN (0.99), Cifar-100 (0.9027) and Cifar-10 (0.9852)). Moreover, our model sets a new record on the Surrey dataset (0.949). The source code of the methods provided in this article is available at https://github.com/leonlha/e2e-3m and http://nguyenhuuphong.me.

arXiv:2010.08376공식 코드 · Papers with Code

Deep and interpretable regression models for ordinal outcomes

Outcomes with a natural order commonly occur in prediction tasks and often the available input data are a mixture of complex data like images and tabular predictors. Deep Learning (DL) models are state-of-the-art for image classification tasks but frequently treat ordinal outcomes as unordered and lack interpretability. In contrast, classical ordinal regression models consider the outcome's order and yield interpretable predictor effects but are limited to tabular data. We present ordinal neural network transformation models (ONTRAMs), which unite DL with classical ordinal regression approaches. ONTRAMs are a special case of transformation models and trade off flexibility and interpretability by additively decomposing the transformation function into terms for image and tabular data using jointly trained neural networks. The performance of the most flexible ONTRAM is by definition equivalent to a standard multi-class DL model trained with cross-entropy while being faster in training when facing ordinal outcomes. Lastly, we discuss how to interpret model components for both tabular and image data on two publicly available datasets.

arXiv:2012.00802

Adversarial Robustness Across Representation Spaces

Adversarial robustness corresponds to the susceptibility of deep neural networks to imperceptible perturbations made at test time. In the context of image tasks, many algorithms have been proposed to make neural networks robust to adversarial perturbations made to the input pixels. These perturbations are typically measured in an $\ell_p$ norm. However, robustness often holds only for the specific attack used for training. In this work we extend the above setting to consider the problem of training of deep neural networks that can be made simultaneously robust to perturbations applied in multiple natural representation spaces. For the case of image data, examples include the standard pixel representation as well as the representation in the discrete cosine transform~(DCT) basis. We design a theoretically sound algorithm with formal guarantees for the above problem. Furthermore, our guarantees also hold when the goal is to require robustness with respect to multiple $\ell_p$ norm based attacks. We then derive an efficient practical implementation and demonstrate the effectiveness of our approach on standard datasets for image classification.

arXiv:2106.15299

Cells are Actors: Social Network Analysis with Classical ML for SOTA Histology Image Classification

Digitization of histology images and the advent of new computational methods, like deep learning, have helped the automatic grading of colorectal adenocarcinoma cancer (CRA). Present automated CRA grading methods, however, usually use tiny image patches and thus fail to integrate the entire tissue micro-architecture for grading purposes. To tackle these challenges, we propose to use a statistical network analysis method to describe the complex structure of the tissue micro-environment by modelling nuclei and their connections as a network. We show that by analyzing only the interactions between the cells in a network, we can extract highly discriminative statistical features for CRA grading. Unlike other deep learning or convolutional graph-based approaches, our method is highly scalable (can be used for cell networks consist of millions of nodes), completely explainable, and computationally inexpensive. We create cell networks on a broad CRC histology image dataset, experiment with our method, and report state-of-the-art performance for the prediction of three-class CRA grading.

arXiv:2109.01486

Studying the Effects of Self-Attention for Medical Image Analysis

When the trained physician interprets medical images, they understand the clinical importance of visual features. By applying cognitive attention, they apply greater focus onto clinically relevant regions while disregarding unnecessary features. The use of computer vision to automate the classification of medical images is widely studied. However, the standard convolutional neural network (CNN) does not necessarily employ subconscious feature relevancy evaluation techniques similar to the trained medical specialist and evaluates features more generally. Self-attention mechanisms enable CNNs to focus more on semantically important regions or aggregated relevant context with long-range dependencies. By using attention, medical image analysis systems can potentially become more robust by focusing on more important clinical feature regions. In this paper, we provide a comprehensive comparison of various state-of-the-art self-attention mechanisms across multiple medical image analysis tasks. Through both quantitative and qualitative evaluations along with a clinical user-centric survey study, we aim to provide a deeper understanding of the effects of self-attention in medical computer vision tasks.

arXiv:2207.04295

Explainable AI (XAI) in Biomedical Signal and Image Processing: Promises and Challenges

Artificial intelligence has become pervasive across disciplines and fields, and biomedical image and signal processing is no exception. The growing and widespread interest on the topic has triggered a vast research activity that is reflected in an exponential research effort. Through study of massive and diverse biomedical data, machine and deep learning models have revolutionized various tasks such as modeling, segmentation, registration, classification and synthesis, outperforming traditional techniques. However, the difficulty in translating the results into biologically/clinically interpretable information is preventing their full exploitation in the field. Explainable AI (XAI) attempts to fill this translational gap by providing means to make the models interpretable and providing explanations. Different solutions have been proposed so far and are gaining increasing interest from the community. This paper aims at providing an overview on XAI in biomedical data processing and points to an upcoming Special Issue on Deep Learning in Biomedical Image and Signal Processing of the IEEE Signal Processing Magazine that is going to appear in March 2022.

arXiv:2305.05511공식 코드 · Papers with Code

Self-supervised dense representation learning for live-cell microscopy with time arrow prediction

State-of-the-art object detection and segmentation methods for microscopy images rely on supervised machine learning, which requires laborious manual annotation of training data. Here we present a self-supervised method based on time arrow prediction pre-training that learns dense image representations from raw, unlabeled live-cell microscopy videos. Our method builds upon the task of predicting the correct order of time-flipped image regions via a single-image feature extractor followed by a time arrow prediction head that operates on the fused features. We show that the resulting dense representations capture inherently time-asymmetric biological processes such as cell divisions on a pixel-level. We furthermore demonstrate the utility of these representations on several live-cell microscopy datasets for detection and segmentation of dividing cells, as well as for cell state classification. Our method outperforms supervised methods, particularly when only limited ground truth annotations are available as is commonly the case in practice. We provide code at https://github.com/weigertlab/tarrow.

arXiv:2405.15661

Exposing Image Classifier Shortcuts with Counterfactual Frequency (CoF) Tables

The rise of deep learning in image classification has brought unprecedented accuracy but also highlighted a key issue: the use of 'shortcuts' by models. Such shortcuts are easy-to-learn patterns from the training data that fail to generalise to new data. Examples include the use of a copyright watermark to recognise horses, snowy background to recognise huskies, or ink markings to detect malignant skin lesions. The explainable AI (XAI) community has suggested using instance-level explanations to detect shortcuts without external data, but this requires the examination of many explanations to confirm the presence of such shortcuts, making it a labour-intensive process. To address these challenges, we introduce Counterfactual Frequency (CoF) tables, a novel approach that aggregates instance-based explanations into global insights, and exposes shortcuts. The aggregation implies the need for some semantic concepts to be used in the explanations, which we solve by labelling the segments of an image. We demonstrate the utility of CoF tables across several datasets, revealing the shortcuts learned from them.