PyTorchKR Papers

“Image Classification”

총 18,403편
태그·기간 등 상세 필터

카테고리 탐색

arXiv:2208.05864

Face Morphing Attacks and Face Image Quality: The Effect of Morphing and the Unsupervised Attack Detection by Quality

Morphing attacks are a form of presentation attacks that gathered increasing attention in recent years. A morphed image can be successfully verified to multiple identities. This operation, therefore, poses serious security issues related to the ability of a travel or identity document to be verified to belong to multiple persons. Previous works touched on the issue of the quality of morphing attack images, however, with the main goal of quantitatively proofing the realistic appearance of the produced morphing attacks. We theorize that the morphing processes might have an effect on both, the perceptual image quality and the image utility in face recognition (FR) when compared to bona fide samples. Towards investigating this theory, this work provides an extensive analysis of the effect of morphing on face image quality, including both general image quality measures and face image utility measures. This analysis is not limited to a single morphing technique, but rather looks at six different morphing techniques and five different data sources using ten different quality measures. This analysis reveals consistent separability between the quality scores of morphing attack and bona fide samples measured by certain quality measures. Our study goes further to build on this effect and investigate the possibility of performing unsupervised morphing attack detection (MAD) based on quality scores. Our study looks intointra and inter-dataset detectability to evaluate the generalizability of such a detection concept on different morphing techniques and bona fide sources. Our final results point out that a set of quality measures, such as MagFace and CNNNIQA, can be used to perform unsupervised and generalized MAD with a correct classification accuracy of over 70%.

arXiv:1712.02976공식 코드 · Papers with Code

Defense against Adversarial Attacks Using High-Level Representation Guided Denoiser

Neural networks are vulnerable to adversarial examples, which poses a threat to their application in security sensitive systems. We propose high-level representation guided denoiser (HGD) as a defense for image classification. Standard denoiser suffers from the error amplification effect, in which small residual adversarial noise is progressively amplified and leads to wrong classifications. HGD overcomes this problem by using a loss function defined as the difference between the target model's outputs activated by the clean image and denoised image. Compared with ensemble adversarial training which is the state-of-the-art defending method on large images, HGD has three advantages. First, with HGD as a defense, the target model is more robust to either white-box or black-box adversarial attacks. Second, HGD can be trained on a small subset of the images and generalizes well to other images and unseen classes. Third, HGD can be transferred to defend models other than the one guiding it. In NIPS competition on defense against adversarial attacks, our HGD solution won the first place and outperformed other models by a large margin.

arXiv:1606.01735

Integrated perception with recurrent multi-task neural networks

Modern discriminative predictors have been shown to match natural intelligences in specific perceptual tasks in image classification, object and part detection, boundary extraction, etc. However, a major advantage that natural intelligences still have is that they work well for "all" perceptual problems together, solving them efficiently and coherently in an "integrated manner". In order to capture some of these advantages in machine perception, we ask two questions: whether deep neural networks can learn universal image representations, useful not only for a single task but for all of them, and how the solutions to the different tasks can be integrated in this framework. We answer by proposing a new architecture, which we call "MultiNet", in which not only deep image features are shared between tasks, but where tasks can interact in a recurrent manner by encoding the results of their analysis in a common shared representation of the data. In this manner, we show that the performance of individual tasks in standard benchmarks can be improved first by sharing features between them and then, more significantly, by integrating their solutions in the common representation.

arXiv:1512.02665

Fine-grained Image Classification by Exploring Bipartite-Graph Labels

Given a food image, can a fine-grained object recognition engine tell "which restaurant which dish" the food belongs to? Such ultra-fine grained image recognition is the key for many applications like search by images, but it is very challenging because it needs to discern subtle difference between classes while dealing with the scarcity of training data. Fortunately, the ultra-fine granularity naturally brings rich relationships among object classes. This paper proposes a novel approach to exploit the rich relationships through bipartite-graph labels (BGL). We show how to model BGL in an overall convolutional neural networks and the resulting system can be optimized through back-propagation. We also show that it is computationally efficient in inference thanks to the bipartite structure. To facilitate the study, we construct a new food benchmark dataset, which consists of 37,885 food images collected from 6 restaurants and totally 975 menus. Experimental results on this new food and three other datasets demonstrates BGL advances previous works in fine-grained object recognition. An online demo is available at http://www.f-zhou.com/fg_demo/.

arXiv:1501.01744

Optimal Radiometric Calibration for Camera-Display Communication

We present a novel method for communicating between a camera and display by embedding and recovering hidden and dynamic information within a displayed image. A handheld camera pointed at the display can receive not only the display image, but also the underlying message. These active scenes are fundamentally different from traditional passive scenes like QR codes because image formation is based on display emittance, not surface reflectance. Detecting and decoding the message requires careful photometric modeling for computational message recovery. Unlike standard watermarking and steganography methods that lie outside the domain of computer vision, our message recovery algorithm uses illumination to optically communicate hidden messages in real world scenes. The key innovation of our approach is an algorithm that performs simultaneous radiometric calibration and message recovery in one convex optimization problem. By modeling the photometry of the system using a camera-display transfer function (CDTF), we derive a physics-based kernel function for support vector machine classification. We demonstrate that our method of optimal online radiometric calibration (OORC) leads to an efficient and robust algorithm for computational messaging between nine commercial cameras and displays.

arXiv:1807.05478

A salt and pepper noise image denoising method based on the generative classification

In this paper, an image denoising algorithm is proposed for salt and pepper noise. First, a generative model is built on a patch as a basic unit and then the algorithm locates the image noise within that patch in order to better describe the patch and obtain better subsequent clustering. Second, the algorithm classifies patches using a generative clustering method, thus providing additional similarity information for noise repair and suppressing the interference of noise, abandoning those categories that consist of a smaller number of patches. Finally, the algorithm builds a non-local switching filter to remove the salt and pepper noise. Simulation results show that the proposed algorithm effectively denoises salt and pepper noise of various densities. It obtains a better visual quality and higher peak signal-to-noise ratio score than several state-of-the-art algorithms. In short, our algorithm uses a noisy patch as the basic unit, a patch clustering method to optimize the repair data set as well as obtain a better denoising effect, and provides a guideline for future denoising and repair methods.

arXiv:1809.00646

Detail Preserving Depth Estimation from a Single Image Using Attention Guided Networks

Convolutional Neural Networks have demonstrated superior performance on single image depth estimation in recent years. These works usually use stacked spatial pooling or strided convolution to get high-level information which are common practices in classification task. However, depth estimation is a dense prediction problem and low-resolution feature maps usually generate blurred depth map which is undesirable in application. In order to produce high quality depth map, say clean and accurate, we propose a network consists of a Dense Feature Extractor (DFE) and a Depth Map Generator (DMG). The DFE combines ResNet and dilated convolutions. It extracts multi-scale information from input image while keeping the feature maps dense. As for DMG, we use attention mechanism to fuse multi-scale features produced in DFE. Our Network is trained end-to-end and does not need any post-processing. Hence, it runs fast and can predict depth map in about 15 fps. Experiment results show that our method is competitive with the state-of-the-art in quantitative evaluation, but can preserve better structural details of the scene depth.

arXiv:2004.03153

Adaptive Multiscale Illumination-Invariant Feature Representation for Undersampled Face Recognition

This paper presents an novel illumination-invariant feature representation approach used to eliminate the varying illumination affection in undersampled face recognition. Firstly, a new illumination level classification technique based on Singular Value Decomposition (SVD) is proposed to judge the illumination level of input image. Secondly, we construct the logarithm edgemaps feature (LEF) based on lambertian model and local near neighbor feature of the face image, applying to local region within multiple scales. Then, the illumination level is referenced to construct the high performance LEF as well realize adaptive fusion for multiple scales LEFs for the face image, performing JLEF-feature. In addition, the constrain operation is used to remove the useless high-frequency interference, disentangling useful facial feature edges and constructing AJLEF-face. Finally, the effects of the our methods and other state-of-the-art algorithms including deep learning methods are tested on Extended Yale B, CMU PIE, AR as well as our Self-build Driver database (SDB). The experimental results demonstrate that the JLEF-feature and AJLEF-face outperform other related approaches for undersampled face recognition under varying illumination.

arXiv:2109.06756

ImUnity: a generalizable VAE-GAN solution for multicenter MR image harmonization

ImUnity is an original deep-learning model designed for efficient and flexible MR image harmonization. A VAE-GAN network, coupled with a confusion module and an optional biological preservation module, uses multiple 2D-slices taken from different anatomical locations in each subject of the training database, as well as image contrast transformations for its self-supervised training. It eventually generates 'corrected' MR images that can be used for various multi-center population studies. Using 3 open source databases (ABIDE, OASIS and SRPBS), which contain MR images from multiple acquisition scanner types or vendors and a large range of subjects ages, we show that ImUnity: (1) outperforms state-of-the-art methods in terms of quality of images generated using traveling subjects; (2) removes sites or scanner biases while improving patients classification; (3) harmonizes data coming from new sites or scanners without the need for an additional fine-tuning and (4) allows the selection of multiple MR reconstructed images according to the desired applications. Tested here on T1-weighted images, ImUnity could be used to harmonize other types of medical images.

arXiv:2206.10969

Single Morphing Attack Detection using Siamese Network and Few-shot Learning

Face morphing attack detection is challenging and presents a concrete and severe threat for face verification systems. Reliable detection mechanisms for such attacks, which have been tested with a robust cross-database protocol and unknown morphing tools still is a research challenge. This paper proposes a framework following the Few-Shot-Learning approach that shares image information based on the siamese network using triplet-semi-hard-loss to tackle the morphing attack detection and boost the clustering classification process. This network compares a bona fide or potentially morphed image with triplets of morphing and bona fide face images. Our results show that this new network cluster the data points, and assigns them to classes in order to obtain a lower equal error rate in a cross-database scenario sharing only small image numbers from an unknown database. Few-shot learning helps to boost the learning process. Experimental results using a cross-datasets trained with FRGCv2 and tested with FERET and the AMSL open-access databases reduced the BPCER10 from 43% to 4.91% using ResNet50 and 5.50% for MobileNetV2.

arXiv:2404.12168

Real-World Efficient Blind Motion Deblurring via Blur Pixel Discretization

As recent advances in mobile camera technology have enabled the capability to capture high-resolution images, such as 4K images, the demand for an efficient deblurring model handling large motion has increased. In this paper, we discover that the image residual errors, i.e., blur-sharp pixel differences, can be grouped into some categories according to their motion blur type and how complex their neighboring pixels are. Inspired by this, we decompose the deblurring (regression) task into blur pixel discretization (pixel-level blur classification) and discrete-to-continuous conversion (regression with blur class map) tasks. Specifically, we generate the discretized image residual errors by identifying the blur pixels and then transform them to a continuous form, which is computationally more efficient than naively solving the original regression problem with continuous values. Here, we found that the discretization result, i.e., blur segmentation map, remarkably exhibits visual similarity with the image residual errors. As a result, our efficient model shows comparable performance to state-of-the-art methods in realistic benchmarks, while our method is up to 10 times computationally more efficient.

arXiv:2406.00947

Cross-Dimensional Medical Self-Supervised Representation Learning Based on a Pseudo-3D Transformation

Medical image analysis suffers from a shortage of data, whether annotated or not. This becomes even more pronounced when it comes to 3D medical images. Self-Supervised Learning (SSL) can partially ease this situation by using unlabeled data. However, most existing SSL methods can only make use of data in a single dimensionality (e.g. 2D or 3D), and are incapable of enlarging the training dataset by using data with differing dimensionalities jointly. In this paper, we propose a new cross-dimensional SSL framework based on a pseudo-3D transformation (CDSSL-P3D), that can leverage both 2D and 3D data for joint pre-training. Specifically, we introduce an image transformation based on the im2col algorithm, which converts 2D images into a format consistent with 3D data. This transformation enables seamless integration of 2D and 3D data, and facilitates cross-dimensional self-supervised learning for 3D medical image analysis. We run extensive experiments on 13 downstream tasks, including 2D and 3D classification and segmentation. The results indicate that our CDSSL-P3D achieves superior performance, outperforming other advanced SSL methods.

arXiv:2611.07431v1cs.CV

TRiST: 의료 영상 분석을 위한 위상 강화 Swin Transformer

TRiST: Topology-Reinforced Swin Transformer for Medical Image Analysis

의료 영상에는 작은 병변부터 장기 경계에 이르는 다양한 규모에서 연결 성분과 루프와 같은 위상적 패턴이 존재한다. 계층적 Vision Transformer가 의료 영상 분석을 발전시켜 왔으나, 이러한 모델은 다중 스케일 표현 전반에 위상을 명시적으로 반영하지 않는다. 본 논문은 위상 정보를 특징 구성, 정제, 어텐션에 통합하는 프레임워크인 Topology-Reinforced Swin Transformer(TRiST)를 제안한다. 첫째, 계층적 다중 스케일 위상 표현 알고리즘은 Swin의 네 가지 고유 단계에서 각각 독립적으로 2D 지속 호몰로지(H0 및 H1)를 계산하여, 미세 스케일 위상을 다운샘플링하는 대신 각 단계의 토큰 격자에 정렬된 기술자를 생성한다. 둘째, 밴드 인식 H0/H1 잔차 정제 모듈은 게이트 잔차 네트워크를 사용하여 짧은, 중간, 긴 지속성 정보를 개별적으로 처리한다. 셋째, 지속성 밴드 점진적 어텐션 편향은 단계별 기술자를 학습된 키 중요도 점수로 변환하여 이동 윈도우 어텐션 로짓을 직접 수정한다. 본 논문은 기반이 되는 Swin 기반 백본을 보존한 채 TRiST를 분류용 Topo-SwinV2-B와 분할용 Topo-Swin-UNet으로 구현한다. 다섯 가지 의료 영상 데이터셋에서의 실험 결과, Swin 기반 베이스라인 대비 일관된 성능 향상을 보였으며 기존 위상 증강 방법과 비교해 경쟁력 있는 성능을 나타냈다.

arXiv:2010.07230공식 코드 · Papers with Code

An Evasion Attack against Stacked Capsule Autoencoder

Capsule network is a type of neural network that uses the spatial relationship between features to classify images. By capturing the poses and relative positions between features, its ability to recognize affine transformation is improved, and it surpasses traditional convolutional neural networks (CNNs) when handling translation, rotation and scaling. The Stacked Capsule Autoencoder (SCAE) is the state-of-the-art capsule network. The SCAE encodes an image as capsules, each of which contains poses of features and their correlations. The encoded contents are then input into the downstream classifier to predict the categories of the images. Existing research mainly focuses on the security of capsule networks with dynamic routing or EM routing, and little attention has been given to the security and robustness of the SCAE. In this paper, we propose an evasion attack against the SCAE. After a perturbation is generated based on the output of the object capsules in the model, it is added to an image to reduce the contribution of the object capsules related to the original category of the image so that the perturbed image will be misclassified. We evaluate the attack using an image classification experiment, and the experimental results indicate that the attack can achieve high success rates and stealthiness. It confirms that the SCAE has a security vulnerability whereby it is possible to craft adversarial samples without changing the original structure of the image to fool the classifiers. We hope that our work will make the community aware of the threat of this attack and raise the attention given to the SCAE's security.

arXiv:2301.02621

Deep leakage from gradients

With the development of artificial intelligence technology, Federated Learning (FL) model has been widely used in many industries for its high efficiency and confidentiality. Some researchers have explored its confidentiality and designed some algorithms to attack training data sets, but these algorithms all have their own limitations. Therefore, most people still believe that local machine learning gradient information is safe and reliable. In this paper, an algorithm based on gradient features is designed to attack the federated learning model in order to attract more attention to the security of federated learning systems. In federated learning system, gradient contains little information compared with the original training data set, but this project intends to restore the original training image data through gradient information. Convolutional Neural Network (CNN) has excellent performance in image processing. Therefore, the federated learning model of this project is equipped with Convolutional Neural Network structure, and the model is trained by using image data sets. The algorithm calculates the virtual gradient by generating virtual image labels. Then the virtual gradient is matched with the real gradient to restore the original image. This attack algorithm is written in Python language, uses cat and dog classification Kaggle data sets, and gradually extends from the full connection layer to the convolution layer, thus improving the universality. At present, the average squared error between the data recovered by this algorithm and the original image information is approximately 5, and the vast majority of images can be completely restored according to the gradient information given, indicating that the gradient of federated learning system is not absolutely safe and reliable.

arXiv:1708.00587

Geometric Convolutional Neural Network for Analyzing Surface-Based Neuroimaging Data

The conventional CNN, widely used for two-dimensional images, however, is not directly applicable to non-regular geometric surface, such as a cortical thickness. We propose Geometric CNN (gCNN) that deals with data representation over a spherical surface and renders pattern recognition in a multi-shell mesh structure. The classification accuracy for sex was significantly higher than that of SVM and image based CNN. It only uses MRI thickness data to classify gender but this method can expand to classify disease from other MRI or fMRI data

arXiv:1706.04382

Shape-Color Differential Moment Invariants under Affine Transformations

We propose the general construction formula of shape-color primitives by using partial differentials of each color channel in this paper. By using all kinds of shape-color primitives, shape-color differential moment invariants can be constructed very easily, which are invariant to the shape affine and color affine transforms. 50 instances of SCDMIs are obtained finally. In experiments, several commonly used color descriptors and SCDMIs are used in image classification and retrieval of color images, respectively. By comparing the experimental results, we find that SCDMIs get better results.

arXiv:1703.07131

Knowledge distillation using unlabeled mismatched images

Current approaches for Knowledge Distillation (KD) either directly use training data or sample from the training data distribution. In this paper, we demonstrate effectiveness of 'mismatched' unlabeled stimulus to perform KD for image classification networks. For illustration, we consider scenarios where this is a complete absence of training data, or mismatched stimulus has to be used for augmenting a small amount of training data. We demonstrate that stimulus complexity is a key factor for distillation's good performance. Our examples include use of various datasets for stimulating MNIST and CIFAR teachers.

arXiv:1703.01053

Skin Lesion Classification using Class Activation Map

We proposed a two stage framework with only one network to analyze skin lesion images, we firstly trained a convolutional network to classify these images, and cropped the import regions which the network has the maximum activation value. In the second stage, we retrained this CNN with the image regions extracted from stage one and output the final probabilities. The two stage framework achieved a mean AUC of 0.857 in ISIC-2017 skin lesion validation set and is 0.04 higher than that of the original inputs, 0.821.

arXiv:1911.04120

An empirical study of the relation between network architecture and complexity

In this preregistration submission, we propose an empirical study of how networks handle changes in complexity of the data. We investigate the effect of network capacity on generalization performance in the face of increasing data complexity. For this, we measure the generalization error for an image classification task where the number of classes steadily increases. We compare a number of modern architectures at different scales in this setting. The methodology, setup, and hypotheses described in this proposal were evaluated by peer review before experiments were conducted.