PyTorchKR Papers

최신 논문

총 12,350편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:1812.03190

Deep-RBF Networks Revisited: Robust Classification with Rejection

One of the main drawbacks of deep neural networks, like many other classifiers, is their vulnerability to adversarial attacks. An important reason for their vulnerability is assigning high confidence to regions with few or even no feature points. By feature points, we mean a nonlinear transformation of the input space extracting a meaningful representation of the input data. On the other hand, deep-RBF networks assign high confidence only to the regions containing enough feature points, but they have been discounted due to the widely-held belief that they have the vanishing gradient problem. In this paper, we revisit the deep-RBF networks by first giving a general formulation for them, and then proposing a family of cost functions thereof inspired by metric learning. In the proposed deep-RBF learning algorithm, the vanishing gradient problem does not occur. We make these networks robust to adversarial attack by adding the reject option to their output layer. Through several experiments on the MNIST dataset, we demonstrate that our proposed method not only achieves significant classification accuracy but is also very resistant to various adversarial attacks.

arXiv:1812.03173

Use Dimensionality Reduction and SVM Methods to Increase the Penetration Rate of Computer Networks

In the world today computer networks have a very important position and most of the urban and national infrastructure as well as organizations are managed by computer networks, therefore, the security of these systems against the planned attacks is of great importance. Therefore, researchers have been trying to find these vulnerabilities so that after identifying ways to penetrate the system, they will provide system protection through preventive or countermeasures. SVM is one of the major algorithms for intrusion detection. In this research, we studied a variety of malware and methods of intrusion detection, provide an efficient method for detecting attacks and utilizing dimension reduction.Thus, we will be able to detect attacks by carefully combining these two algorithms and pre-processes that are performed before the two on the input data. The main question raised is how we can identify attacks on computer networks with the above-mentioned method. In anomalies diagnostic method, by identifying behavior as a normal behavior for the user, the host, or the whole system, any deviation from this behavior is considered as an abnormal behavior, which can be a potential occurrence of an attack. The network intrusion detection system is used by anomaly detection method that uses the SVM algorithm for classification and SVD to reduce the size. Steps of the proposed method include pre-processing of the data set, feature selection, support vector machine, and evaluation.The NSL-KDD data set has been used to teach and test the proposed model. In this study, we inferred the intrusion detection using the SVM algorithm for classification and SVD for diminishing dimensions with no classification algorithm.Also the KNN algorithm has been compared in situations with and without diminishing dimensions,the results have shown that the proposed method has a better performance than comparable methods.

arXiv:1812.03982공식 코드 · Papers with Code

SlowFast Networks for Video Recognition

We present SlowFast networks for video recognition. Our model involves (i) a Slow pathway, operating at low frame rate, to capture spatial semantics, and (ii) a Fast pathway, operating at high frame rate, to capture motion at fine temporal resolution. The Fast pathway can be made very lightweight by reducing its channel capacity, yet can learn useful temporal information for video recognition. Our models achieve strong performance for both action classification and detection in video, and large improvements are pin-pointed as contributions by our SlowFast concept. We report state-of-the-art accuracy on major video recognition benchmarks, Kinetics, Charades and AVA. Code has been made available at: https://github.com/facebookresearch/SlowFast

arXiv:1812.03944

Data Fine-tuning

In real-world applications, commercial off-the-shelf systems are utilized for performing automated facial analysis including face recognition, emotion recognition, and attribute prediction. However, a majority of these commercial systems act as black boxes due to the inaccessibility of the model parameters which makes it challenging to fine-tune the models for specific applications. Stimulated by the advances in adversarial perturbations, this research proposes the concept of Data Fine-tuning to improve the classification accuracy of a given model without changing the parameters of the model. This is accomplished by modeling it as data (image) perturbation problem. A small amount of "noise" is added to the input with the objective of minimizing the classification loss without affecting the (visual) appearance. Experiments performed on three publicly available datasets LFW, CelebA, and MUCT, demonstrate the effectiveness of the proposed concept.

arXiv:1812.03928공식 코드 · Papers with Code

Learning Representations of Sets through Optimized Permutations

Representations of sets are challenging to learn because operations on sets should be permutation-invariant. To this end, we propose a Permutation-Optimisation module that learns how to permute a set end-to-end. The permuted set can be further processed to learn a permutation-invariant representation of that set, avoiding a bottleneck in traditional set models. We demonstrate our model's ability to learn permutations and set representations with either explicit or implicit supervision on four datasets, on which we achieve state-of-the-art results: number sorting, image mosaics, classification from image mosaics, and visual question answering.

arXiv:1812.03880

Automatic Classification of Knee Rehabilitation Exercises Using a Single Inertial Sensor: a Case Study

Inertial measurement units have the ability to accurately record the acceleration and angular velocity of human limb segments during discrete joint movements. These movements are commonly used in exercise rehabilitation programmes following orthopaedic surgery such as total knee replacement. This provides the potential for a biofeedback system with data mining technique for patients undertaking exercises at home without physician supervision. We propose to use machine learning techniques to automatically analyse inertial measurement unit data collected during these exercises, and then assess whether each repetition of the exercise was executed correctly or not. Our approach consists of two main phases: signal segmentation, and segment classification. Accurate pre-processing and feature extraction are paramount topics in order for the technique to work. In this paper, we present a classification method for unsupervised rehabilitation exercises, based on a segmentation process that extracts repetitions from a longer signal activity. The results obtained from experimental datasets of both clinical and healthy subjects, for a set of 4 knee exercises commonly used in rehabilitation, are very promising.

arXiv:1812.03664공식 코드 · Papers with Code

Few-Shot Learning via Embedding Adaptation with Set-to-Set Functions

Learning with limited data is a key challenge for visual recognition. Many few-shot learning methods address this challenge by learning an instance embedding function from seen classes and apply the function to instances from unseen classes with limited labels. This style of transfer learning is task-agnostic: the embedding function is not learned optimally discriminative with respect to the unseen classes, where discerning among them leads to the target task. In this paper, we propose a novel approach to adapt the instance embeddings to the target classification task with a set-to-set function, yielding embeddings that are task-specific and are discriminative. We empirically investigated various instantiations of such set-to-set functions and observed the Transformer is most effective -- as it naturally satisfies key properties of our desired model. We denote this model as FEAT (few-shot embedding adaptation w/ Transformer) and validate it on both the standard few-shot classification benchmark and four extended few-shot learning settings with essential use cases, i.e., cross-domain, transductive, generalized few-shot learning, and low-shot learning. It archived consistent improvements over baseline models as well as previous methods and established the new state-of-the-art results on two benchmarks.

arXiv:1812.03599

Fast convergence rates of deep neural networks for classification

We derive the fast convergence rates of a deep neural network (DNN) classifier with the rectified linear unit (ReLU) activation function learned using the hinge loss. We consider three cases for a true model: (1) a smooth decision boundary, (2) smooth conditional class probability, and (3) the margin condition (i.e., the probability of inputs near the decision boundary is small). We show that the DNN classifier learned using the hinge loss achieves fast rate convergences for all three cases provided that the architecture (i.e., the number of layers, number of nodes and sparsity). is carefully selected. An important implication is that DNN architectures are very flexible for use in various cases without much modification. In addition, we consider a DNN classifier learned by minimizing the cross-entropy, and show that the DNN classifier achieves a fast convergence rate under the condition that the conditional class probabilities of most data are sufficiently close to either 1 or zero. This assumption is not unusual for image recognition because human beings are extremely good at recognizing most images. To confirm our theoretical explanation, we present the results of a small numerical study conducted to compare the hinge loss and cross-entropy.

arXiv:1812.03527

A Deep Multi-task Learning Approach to Skin Lesion Classification

Skin lesion identification is a key step toward dermatological diagnosis. When describing a skin lesion, it is very important to note its body site distribution as many skin diseases commonly affect particular parts of the body. To exploit the correlation between skin lesions and their body site distributions, in this study, we investigate the possibility of improving skin lesion classification using the additional context information provided by body location. Specifically, we build a deep multi-task learning (MTL) framework to jointly optimize skin lesion classification and body location classification (the latter is used as an inductive bias). Our MTL framework uses the state-of-the-art ImageNet pretrained model with specialized loss functions for the two related tasks. Our experiments show that the proposed MTL based method performs more robustly than its standalone (single-task) counterpart.

arXiv:1812.03520

Skin Disease Classification versus Skin Lesion Characterization: Achieving Robust Diagnosis using Multi-label Deep Neural Networks

In this study, we investigate what a practically useful approach is in order to achieve robust skin disease diagnosis. A direct approach is to target the ground truth diagnosis labels, while an alternative approach instead focuses on determining skin lesion characteristics that are more visually consistent and discernible. We argue that, for computer-aided skin disease diagnosis, it is both more realistic and more useful that lesion type tags should be considered as the target of an automated diagnosis system such that the system can first achieve a high accuracy in describing skin lesions, and in turn facilitate disease diagnosis using lesion characteristics in conjunction with other evidence. To further meet such an objective, we employ convolutional neural networks (CNNs) for both the disease-targeted and lesion-targeted classifications. We have collected a large-scale and diverse dataset of 75,665 skin disease images from six publicly available dermatology atlantes. Then we train and compare both disease-targeted and lesion-targeted classifiers, respectively. For disease-targeted classification, only 27.6% top-1 accuracy and 57.9% top-5 accuracy are achieved with a mean average precision (mAP) of 0.42. In contrast, for lesion-targeted classification, we can achieve a much higher mAP of 0.70.

arXiv:1812.03519

Deep-Net: Deep Neural Network for Cyber Security Use Cases

Deep neural networks (DNNs) have witnessed as a powerful approach in this year by solving long-standing Artificial intelligence (AI) supervised and unsupervised tasks exists in natural language processing, speech processing, computer vision and others. In this paper, we attempt to apply DNNs on three different cyber security use cases: Android malware classification, incident detection and fraud detection. The data set of each use case contains real known benign and malicious activities samples. The efficient network architecture for DNN is chosen by conducting various trails of experiments for network parameters and network structures. The experiments of such chosen efficient configurations of DNNs are run up to 1000 epochs with learning rate set in the range [0.01-0.5]. Experiments of DNN performed well in comparison to the classical machine learning algorithms in all cases of experiments of cyber security use cases. This is due to the fact that DNNs implicitly extract and build better features, identifies the characteristics of the data that lead to better accuracy. The best accuracy obtained by DNN and XGBoost on Android malware classification 0.940 and 0.741, incident detection 1.00 and 0.997 fraud detection 0.972 and 0.916 respectively.

arXiv:1812.03411공식 코드 · Papers with Code

Feature Denoising for Improving Adversarial Robustness

Adversarial attacks to image classification systems present challenges to convolutional networks and opportunities for understanding them. This study suggests that adversarial perturbations on images lead to noise in the features constructed by these networks. Motivated by this observation, we develop new network architectures that increase adversarial robustness by performing feature denoising. Specifically, our networks contain blocks that denoise the features using non-local means or other filters; the entire networks are trained end-to-end. When combined with adversarial training, our feature denoising networks substantially improve the state-of-the-art in adversarial robustness in both white-box and black-box attack settings. On ImageNet, under 10-iteration PGD white-box attacks where prior art has 27.9% accuracy, our method achieves 55.7%; even under extreme 2000-iteration PGD white-box attacks, our method secures 42.6% accuracy. Our method was ranked first in Competition on Adversarial Attacks and Defenses (CAAD) 2018 --- it achieved 50.6% classification accuracy on a secret, ImageNet-like test dataset against 48 unknown attackers, surpassing the runner-up approach by ~10%. Code is available at https://github.com/facebookresearch/ImageNet-Adversarial-Training.

arXiv:1812.01941

Anomaly Detection for Network Connection Logs

We leverage a streaming architecture based on ELK, Spark and Hadoop in order to collect, store, and analyse database connection logs in near real-time. The proposed system investigates outliers using unsupervised learning; widely adopted clustering and classification algorithms for log data, highlighting the subtle variances in each model by visualisation of outliers. Arriving at a novel solution to evaluate untagged, unfiltered connection logs, we propose an approach that can be extrapolated to a generalised system of analysing connection logs across a large infrastructure comprising thousands of individual nodes and generating hundreds of lines in logs per second.

arXiv:1812.03170공식 코드 · Papers with Code

Variational Saccading: Efficient Inference for Large Resolution Images

Image classification with deep neural networks is typically restricted to images of small dimensionality such as 224 x 244 in Resnet models [24]. This limitation excludes the 4000 x 3000 dimensional images that are taken by modern smartphone cameras and smart devices. In this work, we aim to mitigate the prohibitive inferential and memory costs of operating in such large dimensional spaces. To sample from the high-resolution original input distribution, we propose using a smaller proxy distribution to learn the co-ordinates that correspond to regions of interest in the high-dimensional space. We introduce a new principled variational lower bound that captures the relationship of the proxy distribution's posterior and the original image's co-ordinate space in a way that maximizes the conditional classification likelihood. We empirically demonstrate on one synthetic benchmark and one real world large resolution DSLR camera image dataset that our method produces comparable results with ~10x faster inference and lower memory consumption than a model that utilizes the entire original input distribution. Finally, we experiment with a more complex setting using mini-maps from Starcraft II [56] to infer the number of characters in a complex 3d-rendered scene. Even in such complicated scenes our model provides strong localization: a feature missing from traditional classification models.

arXiv:1812.03123

Deep Variational Transfer: Transfer Learning through Semi-supervised Deep Generative Models

In real-world applications, it is often expensive and time-consuming to obtain labeled examples. In such cases, knowledge transfer from related domains, where labels are abundant, could greatly reduce the need for extensive labeling efforts. In this scenario, transfer learning comes in hand. In this paper, we propose Deep Variational Transfer (DVT), a variational autoencoder that transfers knowledge across domains using a shared latent Gaussian mixture model. Thanks to the combination of a semi-supervised ELBO and parameters sharing across domains, we are able to simultaneously: (i) align all supervised examples of the same class into the same latent Gaussian Mixture component, independently from their domain; (ii) predict the class of unsupervised examples from different domains and use them to better model the occurring shifts. We perform tests on MNIST and USPS digits datasets, showing DVT's ability to perform transfer learning across heterogeneous datasets. Additionally, we present DVT's top classification performances on the MNIST semi-supervised learning challenge. We further validate DVT on a astronomical datasets. DVT achieves states-of-the-art classification performances, transferring knowledge across real stars surveys datasets, EROS, MACHO and HiTS, . In the worst performance, we double the achieved F1-score for rare classes. These experiments show DVT's ability to tackle all major challenges posed by transfer learning: different covariate distributions, different and highly imbalanced class distributions and different feature spaces.

arXiv:1812.02975

ShuffleNASNets: Efficient CNN models through modified Efficient Neural Architecture Search

Neural network architectures found by sophistic search algorithms achieve strikingly good test performance, surpassing most human-crafted network models by significant margins. Although computationally efficient, their design is often very complex, impairing execution speed. Additionally, finding models outside of the search space is not possible by design. While our space is still limited, we implement undiscoverable expert knowledge into the economic search algorithm Efficient Neural Architecture Search (ENAS), guided by the design principles and architecture of ShuffleNet V2. While maintaining baseline-like 2.85% test error on CIFAR-10, our ShuffleNASNets are significantly less complex, require fewer parameters, and are two times faster than the ENAS baseline in a classification task. These models also scale well to a low parameter space, achieving less than 5% test error with little regularization and only 236K parameters.

arXiv:1812.02956

LNEMLC: Label Network Embeddings for Multi-Label Classification

Multi-label classification aims to classify instances with discrete non-exclusive labels. Most approaches on multi-label classification focus on effective adaptation or transformation of existing binary and multi-class learning approaches but fail in modelling the joint probability of labels or do not preserve generalization abilities for unseen label combinations. To address these issues we propose a new multi-label classification scheme, LNEMLC - Label Network Embedding for Multi-Label Classification, that embeds the label network and uses it to extend input space in learning and inference of any base multi-label classifier. The approach allows capturing of labels' joint probability at low computational complexity providing results comparable to the best methods reported in the literature. We demonstrate how the method reveals statistically significant improvements over the simple kNN baseline classifier. We also provide hints for selecting the robust configuration that works satisfactorily across data domains.

arXiv:1812.02937

Optimizing speed/accuracy trade-off for person re-identification via knowledge distillation

Finding a person across a camera network plays an important role in video surveillance. For a real-world person re-identification application, in order to guarantee an optimal time response, it is crucial to find the balance between accuracy and speed. We analyse this trade-off, comparing a classical method, that comprises hand-crafted feature description and metric learning, in particular, LOMO and XQDA, to deep learning based techniques, using image classification networks, ResNet and MobileNets. Additionally, we propose and analyse network distillation as a learning strategy to reduce the computational cost of the deep learning approach at test time. We evaluate both methods on the Market-1501 and DukeMTMC-reID large-scale datasets, showing that distillation helps reducing the computational cost at inference time while even increasing the accuracy performance.

arXiv:1812.02934

Local Distribution in Neighborhood for Classification

The k-nearest-neighbor method performs classification tasks for a query sample based on the information contained in its neighborhood. Previous studies into the k-nearest-neighbor algorithm usually achieved the decision value for a class by combining the support of each sample in the neighborhood. They have generally considered the nearest neighbors separately, and potentially integral neighborhood information important for classification was lost, e.g. the distribution information. This article proposes a novel local learning method that organizes the information in the neighborhood through local distribution. In the proposed method, additional distribution information in the neighborhood is estimated and then organized; the classification decision is made based on maximum posterior probability which is estimated from the local distribution in the neighborhood. Additionally, based on the local distribution, we generate a generalized local classification form that can be effectively applied to various datasets through tuning the parameters. We use both synthetic and real datasets to evaluate the classification performance of the proposed method; the experimental results demonstrate the dimensional scalability, efficiency, effectiveness and robustness of the proposed method compared to some other state-of-the-art classifiers. The results indicate that the proposed method is effective and promising in a broad range of domains.

arXiv:1812.02891

Adversarial Defense of Image Classification Using a Variational Auto-Encoder

Deep neural networks are known to be vulnerable to adversarial attacks. This exposes them to potential exploits in security-sensitive applications and highlights their lack of robustness. This paper uses a variational auto-encoder (VAE) to defend against adversarial attacks for image classification tasks. This VAE defense has a few nice properties: (1) it is quite flexible and its use of randomness makes it harder to attack; (2) it can learn disentangled representations that prevent blurry reconstruction; and (3) a patch-wise VAE defense strategy is used that does not require retraining for different size images. For moderate to severe attacks, this system outperforms or closely matches the performance of JPEG compression, with the best quality parameter. It also has more flexibility and potential for improvement via training.