PyTorchKR Papers

“Image Classification”

총 18,403편
태그·기간 등 상세 필터

카테고리 탐색

arXiv:1406.7075

Adaptive texture energy measure method

Recent developments in image quality, data storage, and computational capacity have heightened the need for texture analysis in image process. To date various methods have been developed and introduced for assessing textures in images. One of the most popular texture analysis methods is the Texture Energy Measure (TEM) and it has been used for detecting edges, levels, waves, spots and ripples by employing predefined TEM masks to images. Despite several success- ful studies, TEM has a number of serious weaknesses in use. The major drawback is; the masks are predefined therefore they cannot be adapted to image. A new method, Adaptive Texture Energy Measure Method (aTEM), was offered to over- come this disadvantage of TEM by using adaptive masks by adjusting the contrast, sharpening and orientation angle of the mask. To assess the applicability of aTEM, it is compared with TEM. The accuracy of the classification of butterfly, flower seed and Brodatz datasets are 0.08, 0.3292 and 0.3343, respectively by TEM and 0.0053, 0.2417 and 0.3153, respectively by aTEM. The results of this study indicate that aTEM is a successful method for texture analysis.

arXiv:2107.00272

A Survey on Graph-Based Deep Learning for Computational Histopathology

With the remarkable success of representation learning for prediction problems, we have witnessed a rapid expansion of the use of machine learning and deep learning for the analysis of digital pathology and biopsy image patches. However, learning over patch-wise features using convolutional neural networks limits the ability of the model to capture global contextual information and comprehensively model tissue composition. The phenotypical and topological distribution of constituent histological entities play a critical role in tissue diagnosis. As such, graph data representations and deep learning have attracted significant attention for encoding tissue representations, and capturing intra- and inter- entity level interactions. In this review, we provide a conceptual grounding for graph analytics in digital pathology, including entity-graph construction and graph architectures, and present their current success for tumor localization and classification, tumor invasion and staging, image retrieval, and survival prediction. We provide an overview of these methods in a systematic manner organized by the graph representation of the input image, scale, and organ on which they operate. We also outline the limitations of existing techniques, and suggest potential future research directions in this domain.

arXiv:2204.02958

LEAD: Self-Supervised Landmark Estimation by Aligning Distributions of Feature Similarity

In this work, we introduce LEAD, an approach to discover landmarks from an unannotated collection of category-specific images. Existing works in self-supervised landmark detection are based on learning dense (pixel-level) feature representations from an image, which are further used to learn landmarks in a semi-supervised manner. While there have been advances in self-supervised learning of image features for instance-level tasks like classification, these methods do not ensure dense equivariant representations. The property of equivariance is of interest for dense prediction tasks like landmark estimation. In this work, we introduce an approach to enhance the learning of dense equivariant representations in a self-supervised fashion. We follow a two-stage training approach: first, we train a network using the BYOL objective which operates at an instance level. The correspondences obtained through this network are further used to train a dense and compact representation of the image using a lightweight network. We show that having such a prior in the feature extractor helps in landmark detection, even under drastically limited number of annotations while also improving generalization across scale variations.

arXiv:2506.13496

Hierarchical Multi-Positive Contrastive Learning for Patent Image Retrieval

Patent images are technical drawings that convey information about a patent's innovation. Patent image retrieval systems aim to search in vast collections and retrieve the most relevant images. Despite recent advances in information retrieval, patent images still pose significant challenges due to their technical intricacies and complex semantic information, requiring efficient fine-tuning for domain adaptation. Current methods neglect patents' hierarchical relationships, such as those defined by the Locarno International Classification (LIC) system, which groups broad categories (e.g., "furnishing") into subclasses (e.g., "seats" and "beds") and further into specific patent designs. In this work, we introduce a hierarchical multi-positive contrastive loss that leverages the LIC's taxonomy to induce such relations in the retrieval process. Our approach assigns multiple positive pairs to each patent image within a batch, with varying similarity scores based on the hierarchical taxonomy. Our experimental analysis with various vision and multimodal models on the DeepPatent2 dataset shows that the proposed method enhances the retrieval results. Notably, our method is effective with low-parameter models, which require fewer computational resources and can be deployed on environments with limited hardware.

arXiv:1910.08336

KerCNNs: biologically inspired lateral connections for classification of corrupted images

The state of the art in many computer vision tasks is represented by Convolutional Neural Networks (CNNs). Although their hierarchical organization and local feature extraction are inspired by the structure of primate visual systems, the lack of lateral connections in such architectures critically distinguishes their analysis from biological object processing. The idea of enriching CNNs with recurrent lateral connections of convolutional type has been put into practice in recent years, in the form of learned recurrent kernels with no geometrical constraints. In the present work, we introduce biologically plausible lateral kernels encoding a notion of correlation between the feedforward filters of a CNN: at each layer, the associated kernel acts as a transition kernel on the space of activations. The lateral kernels are defined in terms of the filters, thus providing a parameter-free approach to assess the geometry of horizontal connections based on the feedforward structure. We then test this new architecture, which we call KerCNN, on a generalization task related to global shape analysis and pattern completion: once trained for performing basic image classification, the network is evaluated on corrupted testing images. The image perturbations examined are designed to undermine the recognition of the images via local features, thus requiring an integration of context information - which in biological vision is critically linked to lateral connectivity. Our KerCNNs turn out to be far more stable than CNNs and recurrent CNNs to such degradations, thus validating this biologically inspired approach to reinforce object recognition under challenging conditions.

arXiv:2011.11058

Investigating Emotion-Color Association in Deep Neural Networks

It has been found that representations learned by Deep Neural Networks (DNNs) correlate very well to neural responses measured in primates' brains and psychological representations exhibited by human similarity judgment. On another hand, past studies have shown that particular colors can be associated with specific emotion arousal in humans. Do deep neural networks also learn this behavior? In this study, we investigate if DNNs can learn implicit associations in stimuli, particularly, an emotion-color association between image stimuli. Our study was conducted in two parts. First, we collected human responses on a forced-choice decision task in which subjects were asked to select a color for a specified emotion-inducing image. Next, we modeled this decision task on neural networks using the similarity between deep representation (extracted using DNNs trained on object classification tasks) of the images and images of colors used in the task. We found that our model showed a fuzzy linear relationship between the two decision probabilities. This results in two interesting findings, 1. The representations learned by deep neural networks can indeed show an emotion-color association 2. The emotion-color association is not just random but involves some cognitive phenomena. Finally, we also show that this method can help us in the emotion classification task, specifically when there are very few examples to train the model. This analysis can be relevant to psychologists studying emotion-color associations and artificial intelligence researchers modeling emotional intelligence in machines or studying representations learned by deep neural networks.

논문 ID: document/9841445

Few-shot Learning with Class-Covariance Metric for Hyperspectral Image Classification

Recently, embedding and metric-based few-shot learning (FSL) has been introduced into hyperspectral image classification (HSIC) and achieved impressive progress. To further enhance the performance with few labeled samples, we in this paper propose a novel FSL framework for HSIC with a class-covariance metric (CMFSL). Overall, the CMFSL learns global class representations for each training episode by interactively using training samples from the base and novel classes, and a synthesis strategy is employed on the novel classes to avoid overfitting. During the meta-training and meta-testing, the class labels are determined directly using the Mahalanobis distance measurement rather than an extra classifier. Benefiting from the task-adapted class-covariance estimations, the CMFSL can construct more flexible decision boundaries than the commonly used Euclidean metric. Additionally, a lightweight cross-scale convolutional network (LXConvNet) consisting of 3-D and 2-D convolutions is designed to thoroughly exploit the spectral-spatial information in the high-frequency and low-frequency scales with low computational complexity. Furthermore, we devise a spectral-prior-based refinement module (SPRM) in the initial stage of feature extraction, which can not only force the network to emphasize the most informative bands while suppressing the useless ones, but also alleviate the effects of the domain shift between the base and novel categories to learn a collaborative embedding mapping. Extensive experiment results on four benchmark data sets demonstrate that the proposed CMFSL can outperform the state-of-the-art methods with few-shot annotated samples.

arXiv:2310.17250공식 코드 · Papers with Code

Model Input-Output Configuration Search with Embedded Feature Selection for Sensor Time-series and Image Classification

Machine learning is a powerful tool for extracting valuable information and making various predictions from diverse datasets. Traditional machine learning algorithms rely on well-defined input and output variables; however, there are scenarios where the separation between the input and output variables and the underlying, associated input and output layers of the model are unknown. Feature Selection (FS) and Neural Architecture Search (NAS) have emerged as promising solutions in such scenarios. This paper proposes MICS-EFS, a Model Input-Output Configuration Search with Embedded Feature Selection. The methodology explores internal dependencies in the complete input parameter space for classification tasks involving both 1D sensor time-series and 2D image data. MICS-EFS employs a modified encoder-decoder model and the Sequential Forward Search (SFS) algorithm, combining input-output configuration search with embedded feature selection. Experimental results demonstrate the superior performance of MICS-EFS compared to other FS algorithms. Across all tested datasets, MICS-EFS delivered an average accuracy improvement of 1.5% over baseline models, with the accuracy gains ranging from 0.5% to 5.9%. Moreover, the algorithm reduced feature dimensionality to just 2-5% of the original data, significantly enhancing computational efficiency. These results highlight the potential of MICS-EFS to improve model accuracy and efficiency in various machine learning tasks. Furthermore, the proposed method has been validated in a real-world industrial application focused on machining processes, underscoring its effectiveness and practicality in addressing complex input-output challenges.

arXiv:2403.10569

Achieving Pareto Optimality using Efficient Parameter Reduction for DNNs in Resource-Constrained Edge Environment

This paper proposes an optimization of an existing Deep Neural Network (DNN) that improves its hardware utilization and facilitates on-device training for resource-constrained edge environments. We implement efficient parameter reduction strategies on Xception that shrink the model size without sacrificing accuracy, thus decreasing memory utilization during training. We evaluate our model in two experiments: Caltech-101 image classification and PCB defect detection and compare its performance against the original Xception and lightweight models, EfficientNetV2B1 and MobileNetV2. The results of the Caltech-101 image classification show that our model has a better test accuracy (76.21%) than Xception (75.89%), uses less memory on average (847.9MB) than Xception (874.6MB), and has faster training and inference times. The lightweight models overfit with EfficientNetV2B1 having a 30.52% test accuracy and MobileNetV2 having a 58.11% test accuracy. Both lightweight models have better memory usage than our model and Xception. On the PCB defect detection, our model has the best test accuracy (90.30%), compared to Xception (88.10%), EfficientNetV2B1 (55.25%), and MobileNetV2 (50.50%). MobileNetV2 has the least average memory usage (849.4MB), followed by our model (865.8MB), then EfficientNetV2B1 (874.8MB), and Xception has the highest (893.6MB). We further experiment with pre-trained weights and observe that memory usage decreases thereby showing the benefits of transfer learning. A Pareto analysis of the models' performance shows that our optimized model architecture satisfies accuracy and low memory utilization objectives.

arXiv:2405.14242

M2ANET: Mobile Malaria Attention Network for efficient classification of plasmodium parasites in blood cells

Malaria is a life-threatening infectious disease caused by Plasmodium parasites, which poses a significant public health challenge worldwide, particularly in tropical and subtropical regions. Timely and accurate detection of malaria parasites in blood cells is crucial for effective treatment and control of the disease. In recent years, deep learning techniques have demonstrated remarkable success in medical image analysis tasks, offering promising avenues for improving diagnostic accuracy, with limited studies on hybrid mobile models due to the complexity of combining two distinct models and the significant memory demand of self-attention mechanism especially for edge devices. In this study, we explore the potential of designing a hybrid mobile model for efficient classification of plasmodium parasites in blood cell images. Therefore, we present M2ANET (Mobile Malaria Attention Network). The model integrates MBConv3 (MobileNetV3 blocks) for efficient capturing of local feature extractions within blood cell images and a modified global-MHSA (multi-head self-attention) mechanism in the latter stages of the network for capturing global context. Through extensive experimentation on benchmark, we demonstrate that M2ANET outperforms some state-of-the-art lightweight and mobile networks in terms of both accuracy and efficiency. Moreover, we discuss the potential implications of M2ANET in advancing malaria diagnosis and treatment, highlighting its suitability for deployment in resource-constrained healthcare settings. The development of M2ANET represents a significant advancement in the pursuit of efficient and accurate malaria detection, with broader implications for medical image analysis and global healthcare initiatives.

arXiv:2502.20277

Explainable, Multi-modal Wound Infection Classification from Images Augmented with Generated Captions

Infections in Diabetic Foot Ulcers (DFUs) can cause severe complications, including tissue death and limb amputation, highlighting the need for accurate, timely diagnosis. Previous machine learning methods have focused on identifying infections by analyzing wound images alone, without utilizing additional metadata such as medical notes. In this study, we aim to improve infection detection by introducing Synthetic Caption Augmented Retrieval for Wound Infection Detection (SCARWID), a novel deep learning framework that leverages synthetic textual descriptions to augment DFU images. SCARWID consists of two components: (1) Wound-BLIP, a Vision-Language Model (VLM) fine-tuned on GPT-4o-generated descriptions to synthesize consistent captions from images; and (2) an Image-Text Fusion module that uses cross-attention to extract cross-modal embeddings from an image and its corresponding Wound-BLIP caption. Infection status is determined by retrieving the top-k similar items from a labeled support set. To enhance the diversity of training data, we utilized a latent diffusion model to generate additional wound images. As a result, SCARWID outperformed state-of-the-art models, achieving average sensitivity, specificity, and accuracy of 0.85, 0.78, and 0.81, respectively, for wound infection classification. Displaying the generated captions alongside the wound images and infection detection results enhances interpretability and trust, enabling nurses to align SCARWID outputs with their medical knowledge. This is particularly valuable when wound notes are unavailable or when assisting novice nurses who may find it difficult to identify visual attributes of wound infection.

arXiv:2607.28538v1cs.CV

ScaFE: LLM 생성 임상 특징 프로그램을 활용한 데이터 효율적인 흉터 분류

ScaFE: Data-Efficient Scar Classification with LLM-Generated Clinical Feature Programs

임상 사진에서 병리적 흉터를 분류하려면 전문가 라벨링 데이터의 부족과 병원 간 상당한 촬영 편차에도 불구하고 켈로이드(keloid)와 비후성 흉터(hypertrophic scar)를 구분해야 합니다. 엔드투엔드(end-to-end) 이미지 모델은 여전히 데이터 의존적이며, 호스팅된 비전-언어 모델(VLM)로 사진을 전송하는 것은 현지 데이터 거버넌스 요건과 상충될 수 있고 재현 및 감사가 어려운 결정을 낳습니다. 본 논문에서는 대규모 언어 모델(LLM)에 이미지를 직접 진단하도록 요구하는 대신, LLM의 임상 지식을 결정론적이고 실행 가능한 특징 프로그램으로 이전하는 ScaFE(Scar Feature Engineering)를 소개합니다. 웹 검색이 가능한 LLM이 임상적 근거를 검색하고 시각적으로 평가 가능한 흉터 속성을 측정하는 프로그램을 합성합니다. 후보 프로그램은 제한된 로컬 환경에서 실행되며, 반복적인 수정 및 개선을 위해 집계된 검증 통계와 특징 수준의 SHAP 요약만 반환되므로 원본 이미지와 환자 수준의 출력값은 로컬에 유지됩니다. 그런 다음 경량 Random Forest가 결과로 도출된 구조화된 표현을 기반으로 작동합니다. 3개 병원의 임상 사진 600장을 대상으로 한 기관 배제 교차 검증(leave-one-site-out evaluation)에서 ScaFE는 81.0%의 기관 매크로 균형 정확도(site-macro balanced accuracy)를 달성하여 가장 강력한 기준선 모델인 BiomedCLIP을 10.0%p 앞섰습니다. 개발 데이터의 10%만 사용했을 때도 ScaFE는 72.0%의 균형 정확도를 유지하며 11.8%p의 우위를 지켰습니다. 또한 반복적인 개선을 통해 실행 가능한 프로그램 비율을 66.7%에서 95.0%로 끌어올렸으며, 최종 특징의 91.7%에 대해 검증된 근거를 확보했습니다. 이러한 결과는 LLM 지식이 직접적인 VLM 결정 대신 로컬에서 감사 가능한 특징 프로그램을 통해 데이터 효율적인 다기관 의료 이미지 분류를 지원할 수 있음을 보여줍니다.

arXiv:2611.23484v1cs.CV

3D 의료 비전-언어 사전학습을 위한 참조 기반 데이터 큐레이션

Reference-Guided Data Curation for 3D Medical Vision-Language Pretraining

여러 기관에서 수집된 대규모 의료 영상 데이터셋으로 3D 비전-언어 모델(VLM)을 학습하는 일은 데이터 품질에 의해 제약된다. 큐레이션되지 않은 방대한 영상-보고서 쌍 풀은 중복, 클래스 불균형, 기관 간 이질성을 특징으로 한다. 본 연구는 3D VLM 사전학습을 개선하기 위한 데이터 큐레이션 문제를 다루며, 오프라인 방식이면서 학습이 필요 없는 참조 기반 데이터 큐레이션 방법을 제안한다. 제안 방법은 소규모의 고품질 참조 집합을 활용하여, 큐레이션되지 않은 대규모 풀의 각 모달리티에 대한 품질 신호를 독립적으로 추출하고, 순위 집계 단계를 통해 이를 하나의 순위로 병합하여 샘플 선택을 안내한다. 두 기관에서 수집한 3D 복부 CT 볼륨-보고서 쌍 35k개와 2D 흉부 X선 영상-보고서 쌍 58k개로 구성된 데이터셋에서 이 방법을 검증한다. 3D CT에서는 참조 기반 데이터 큐레이션 전략을 사용한 큐레이션 모델이 최신 기준 모델(Merlin)의 제로샷 질환 분류 성능과 동등한 성능을 보이면서도, 학습 데이터는 $3.3\times$ 적게, 참조 집합은 5k개 샘플만 사용한다. 2D 흉부 X선에서는 제안 방법으로 얻은 큐레이션 모델이 전체 풀로 학습한 모델의 제로샷 질환 분류 성능과 같거나 이를 능가하면서, 학습 데이터는 최대 $10\times$ 적게, 참조 집합은 74개 샘플만 사용한다. 절제 실험 결과, 선택은 참조 집합을 재표집하더라도 안정적이고, 더 작은 참조 집합도 무작위 샘플링보다 우수하며, 레이블 재균형은 성능을 향상시키지 않는 것으로 나타났다. 재현성을 지원하기 위해 코드를 공개한다.

arXiv:1605.09584

A Sparse Representation of Complete Local Binary Pattern Histogram for Human Face Recognition

Human face recognition has been a long standing problem in computer vision and pattern recognition. Facial analysis can be viewed as a two-fold problem, namely (i) facial representation, and (ii) classification. So far, many face representations have been proposed, a well-known method is the Local Binary Pattern (LBP), which has witnessed a growing interest. In this respect, we treat in this paper the issues of face representation as well as classification in a novel manner. On the one hand, we use a variant to LBP, so-called Complete Local Binary Pattern (CLBP), which differs from the basic LBP by coding a given local region using a given central pixel and Sing_ Magnitude difference. Subsequently, most of LBPbased descriptors use a fixed grid to code a given facial image, which technique is, in most cases, not robust to pose variation and misalignment. To cope with such issue, a representative Multi-Resolution Histogram (MH) decomposition is adopted in our work. On the other hand, having the histograms of the considered images extracted, we exploit their sparsity to construct a so-called Sparse Representation Classifier (SRC) for further face classification. Experimental results have been conducted on ORL face database, and pointed out the superiority of our scheme over other popular state-of-the-art techniques.

arXiv:1811.11375

Instance-level Sketch-based Retrieval by Deep Triplet Classification Siamese Network

Sketch has been employed as an effective communicative tool to express the abstract and intuitive meanings of object. Recognizing the free-hand sketch drawing is extremely useful in many real-world applications. While content-based sketch recognition has been studied for several decades, the instance-level Sketch-Based Image Retrieval (SBIR) tasks have attracted significant research attention recently. The existing datasets such as QMUL-Chair and QMUL-Shoe, focus on the retrieval tasks of chairs and shoes. However, there are several key limitations in previous instance-level SBIR works. The state-of-the-art works have to heavily rely on the pre-training process, quality of edge maps, multi-cropping testing strategy, and augmenting sketch images. To efficiently solve the instance-level SBIR, we propose a new Deep Triplet Classification Siamese Network (DeepTCNet) which employs DenseNet-169 as the basic feature extractor and is optimized by the triplet loss and classification loss. Critically, our proposed DeepTCNet can break the limitations existed in previous works. The extensive experiments on five benchmark sketch datasets validate the effectiveness of the proposed model. Additionally, to study the tasks of sketch-based hairstyle retrieval, this paper contributes a new instance-level photo-sketch dataset - Hairstyle Photo-Sketch dataset, which is composed of 3600 sketches and photos, and 2400 sketch-photo pairs.

arXiv:1904.12724공식 코드 · Papers with Code

Solo or Ensemble? Choosing a CNN Architecture for Melanoma Classification

Convolutional neural networks (CNNs) deliver exceptional results for computer vision, including medical image analysis. With the growing number of available architectures, picking one over another is far from obvious. Existing art suggests that, when performing transfer learning, the performance of CNN architectures on ImageNet correlates strongly with their performance on target tasks. We evaluate that claim for melanoma classification, over 9 CNNs architectures, in 5 sets of splits created on the ISIC Challenge 2017 dataset, and 3 repeated measures, resulting in 135 models. The correlations we found were, to begin with, much smaller than those reported by existing art, and disappeared altogether when we considered only the top-performing networks: uncontrolled nuisances (i.e., splits and randomness) overcome any of the analyzed factors. Whenever possible, the best approach for melanoma classification is still to create ensembles of multiple models. We compared two choices for selecting which models to ensemble: picking them at random (among a pool of high-quality ones) vs. using the validation set to determine which ones to pick first. For small ensembles, we found a slight advantage on the second approach but found that random choice was also competitive. Although our aim in this paper was not to maximize performance, we easily reached AUCs comparable to the first place on the ISIC Challenge 2017.

arXiv:2004.03188

Increasing the Inference and Learning Speed of Tsetlin Machines with Clause Indexing

The Tsetlin Machine (TM) is a machine learning algorithm founded on the classical Tsetlin Automaton (TA) and game theory. It further leverages frequent pattern mining and resource allocation principles to extract common patterns in the data, rather than relying on minimizing output error, which is prone to overfitting. Unlike the intertwined nature of pattern representation in neural networks, a TM decomposes problems into self-contained patterns, represented as conjunctive clauses. The clause outputs, in turn, are combined into a classification decision through summation and thresholding, akin to a logistic regression function, however, with binary weights and a unit step output function. In this paper, we exploit this hierarchical structure by introducing a novel algorithm that avoids evaluating the clauses exhaustively. Instead we use a simple look-up table that indexes the clauses on the features that falsify them. In this manner, we can quickly evaluate a large number of clauses through falsification, simply by iterating through the features and using the look-up table to eliminate those clauses that are falsified. The look-up table is further structured so that it facilitates constant time updating, thus supporting use also during learning. We report up to 15 times faster classification and three times faster learning on MNIST and Fashion-MNIST image classification, and IMDb sentiment analysis.

arXiv:2104.00222

Embedded Self-Distillation in Compact Multi-Branch Ensemble Network for Remote Sensing Scene Classification

Remote sensing (RS) image scene classification task faces many challenges due to the interference from different characteristics of different geographical elements. To solve this problem, we propose a multi-branch ensemble network to enhance the feature representation ability by fusing features in final output logits and intermediate feature maps. However, simply adding branches will increase the complexity of models and decline the inference efficiency. On this issue, we embed self-distillation (SD) method to transfer knowledge from ensemble network to main-branch in it. Through optimizing with SD, main-branch will have close performance as ensemble network. During inference, we can cut other branches to simplify the whole model. In this paper, we first design compact multi-branch ensemble network, which can be trained in an end-to-end manner. Then, we insert SD method on output logits and feature maps. Compared to previous methods, our proposed architecture (ESD-MBENet) performs strongly on classification accuracy with compact design. Extensive experiments are applied on three benchmark RS datasets AID, NWPU-RESISC45 and UC-Merced with three classic baseline models, VGG16, ResNet50 and DenseNet121. Results prove that our proposed ESD-MBENet can achieve better accuracy than previous state-of-the-art (SOTA) complex models. Moreover, abundant visualization analysis make our method more convincing and interpretable.

arXiv:2109.08020공식 코드 · Papers with Code

The pitfalls of using open data to develop deep learning solutions for COVID-19 detection in chest X-rays

Since the emergence of COVID-19, deep learning models have been developed to identify COVID-19 from chest X-rays. With little to no direct access to hospital data, the AI community relies heavily on public data comprising numerous data sources. Model performance results have been exceptional when training and testing on open-source data, surpassing the reported capabilities of AI in pneumonia-detection prior to the COVID-19 outbreak. In this study impactful models are trained on a widely used open-source data and tested on an external test set and a hospital dataset, for the task of classifying chest X-rays into one of three classes: COVID-19, non-COVID pneumonia and no-pneumonia. Classification performance of the models investigated is evaluated through ROC curves, confusion matrices and standard classification metrics. Explainability modules are implemented to explore the image features most important to classification. Data analysis and model evaluations show that the popular open-source dataset COVIDx is not representative of the real clinical problem and that results from testing on this are inflated. Dependence on open-source data can leave models vulnerable to bias and confounding variables, requiring careful analysis to develop clinically useful/viable AI tools for COVID-19 detection in chest X-rays.

arXiv:2109.11249공식 코드 · Papers with Code

FooBaR: Fault Fooling Backdoor Attack on Neural Network Training

Neural network implementations are known to be vulnerable to physical attack vectors such as fault injection attacks. As of now, these attacks were only utilized during the inference phase with the intention to cause a misclassification. In this work, we explore a novel attack paradigm by injecting faults during the training phase of a neural network in a way that the resulting network can be attacked during deployment without the necessity of further faulting. In particular, we discuss attacks against ReLU activation functions that make it possible to generate a family of malicious inputs, which are called fooling inputs, to be used at inference time to induce controlled misclassifications. Such malicious inputs are obtained by mathematically solving a system of linear equations that would cause a particular behaviour on the attacked activation functions, similar to the one induced in training through faulting. We call such attacks fooling backdoors as the fault attacks at the training phase inject backdoors into the network that allow an attacker to produce fooling inputs. We evaluate our approach against multi-layer perceptron networks and convolutional networks on a popular image classification task obtaining high attack success rates (from 60% to 100%) and high classification confidence when as little as 25 neurons are attacked while preserving high accuracy on the originally intended classification task.