PyTorchKR Papers

최신 논문

총 10,163편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2005.01433

Automated eye disease classification method from anterior eye image using anatomical structure focused image classification technique

This paper presents an automated classification method of infective and non-infective diseases from anterior eye images. Treatments for cases of infective and non-infective diseases are different. Distinguishing them from anterior eye images is important to decide a treatment plan. Ophthalmologists distinguish them empirically. Quantitative classification of them based on computer assistance is necessary. We propose an automated classification method of anterior eye images into cases of infective or non-infective disease. Anterior eye images have large variations of the eye position and brightness of illumination. This makes the classification difficult. If we focus on the cornea, positions of opacified areas in the corneas are different between cases of the infective and non-infective diseases. Therefore, we solve the anterior eye image classification task by using an object detection approach targeting the cornea. This approach can be said as "anatomical structure focused image classification". We use the YOLOv3 object detection method to detect corneas of infective disease and corneas of non-infective disease. The detection result is used to define a classification result of a image. In our experiments using anterior eye images, 88.3% of images were correctly classified by the proposed method.

arXiv:2005.01427공식 코드 · Papers with Code

LIMEtree: Consistent and Faithful Multi-class Explanations

Explainable artificial intelligence provides tools to better understand predictive models and their decisions, but many such methods are limited to producing insights with respect to a single class. When generating explanations for several classes, reasoning over them to obtain a complete view may be difficult since they can present competing or contradictory evidence. To address this challenge we introduce the novel paradigm of multi-class explanations. We outline the theory behind such techniques and propose a local surrogate model based on multi-output regression trees -- called LIMEtree -- that offers faithful and consistent explanations of multiple classes for individual predictions while being post-hoc, model-agnostic and data-universal. On top of strong fidelity guarantees, our implementation delivers a range of diverse explanation types, including counterfactual statements favoured in the literature. We evaluate our algorithm with respect to explainability desiderata, through quantitative experiments and via a pilot user study, on image and tabular data classification tasks, comparing it to LIME, which is a state-of-the-art surrogate explainer. Our contributions demonstrate the benefits of multi-class explanations and wide-ranging advantages of our method across a diverse set of scenarios.

arXiv:2005.00288

Distilling Spikes: Knowledge Distillation in Spiking Neural Networks

Spiking Neural Networks (SNN) are energy-efficient computing architectures that exchange spikes for processing information, unlike classical Artificial Neural Networks (ANN). Due to this, SNNs are better suited for real-life deployments. However, similar to ANNs, SNNs also benefit from deeper architectures to obtain improved performance. Furthermore, like the deep ANNs, the memory, compute and power requirements of SNNs also increase with model size, and model compression becomes a necessity. Knowledge distillation is a model compression technique that enables transferring the learning of a large machine learning model to a smaller model with minimal loss in performance. In this paper, we propose techniques for knowledge distillation in spiking neural networks for the task of image classification. We present ways to distill spikes from a larger SNN, also called the teacher network, to a smaller one, also called the student network, while minimally impacting the classification accuracy. We demonstrate the effectiveness of the proposed method with detailed experiments on three standard datasets while proposing novel distillation methodologies and loss functions. We also present a multi-stage knowledge distillation technique for SNNs using an intermediate network to obtain higher performance from the student network. Our approach is expected to open up new avenues for deploying high performing large SNN models on resource-constrained hardware platforms.

arXiv:2005.00116

Sequence Information Channel Concatenation for Improving Camera Trap Image Burst Classification

Camera Traps are extensively used to observe wildlife in their natural habitat without disturbing the ecosystem. This could help in the early detection of natural or human threats to animals, and help towards ecological conservation. Currently, a massive number of such camera traps have been deployed at various ecological conservation areas around the world, collecting data for decades, thereby requiring automation to detect images containing animals. Existing systems perform classification to detect if images contain animals by considering a single image. However, due to challenging scenes with animals camouflaged in their natural habitat, it sometimes becomes difficult to identify the presence of animals from merely a single image. We hypothesize that a short burst of images instead of a single image, assuming that the animal moves, makes it much easier for a human as well as a machine to detect the presence of animals. In this work, we explore a variety of approaches, and measure the impact of using short image sequences (burst of 3 images) on improving the camera trap image classification. We show that concatenating masks containing sequence information and the images from the 3-image-burst across channels, improves the ROC AUC by 20% on a test-set from unseen camera-sites, as compared to an equivalent model that learns from a single image.

논문 ID: publication/3407327

Fast Gradient Descent Algorithm for Image Classification with Neural Networks

Any optimization of gradient descent methods involves selecting a learning rate. Tuning the learning rate can quickly become repetitive with deeper models of image classification, does not necessarily lead to optimal convergence. We proposed in this paper, a modification of the gradient descent algorithm in which the Nestrove step is added, and the learning rate is update in each epoch. Instead, we learn learning rate itself, either by Armijo rule, or by control step. Our algorithm called fast gradient descent (FGD) for solving image classification with neural networks problems, the quadratic convergence rate o(k^2) of FGD algorithm are proved. FGD algorithm are applicate to a MNIST dataset. The numerical experiment, show that our approach FGD algorithm is faster than gradient descent algorithms.

arXiv:2004.14798공식 코드 · Papers with Code

RAIN: A Simple Approach for Robust and Accurate Image Classification Networks

It has been shown that the majority of existing adversarial defense methods achieve robustness at the cost of sacrificing prediction accuracy. The undesirable severe drop in accuracy adversely affects the reliability of machine learning algorithms and prohibits their deployment in realistic applications. This paper aims to address this dilemma by proposing a novel preprocessing framework, which we term Robust and Accurate Image classificatioN(RAIN), to improve the robustness of given CNN classifiers and, at the same time, preserve their high prediction accuracies. RAIN introduces a new randomization-enhancement scheme. It applies randomization over inputs to break the ties between the model forward prediction path and the backward gradient path, thus improving the model robustness. However, similar to existing preprocessing-based methods, the randomized process will degrade the prediction accuracy. To understand why this is the case, we compare the difference between original and processed images, and find it is the loss of high-frequency components in the input image that leads to accuracy drop of the classifier. Based on this finding, RAIN enhances the input's high-frequency details to retain the CNN's high prediction accuracy. Concretely, RAIN consists of two novel randomization modules: randomized small circular shift (RdmSCS) and randomized down-upsampling (RdmDU). The RdmDU module randomly downsamples the input image, and then the RdmSCS module circularly shifts the input image along a randomly chosen direction by a small but random number of pixels. Finally, the RdmDU module performs upsampling with a detail-enhancement model, such as deep super-resolution networks. We conduct extensive experiments on the STL10 and ImageNet datasets to verify the effectiveness of RAIN against various types of adversarial attacks.

arXiv:2004.14477공식 코드 · Papers with Code

Packet2Vec: Utilizing Word2Vec for Feature Extraction in Packet Data

One of deep learning's attractive benefits is the ability to automatically extract relevant features for a target problem from largely raw data, instead of utilizing human engineered and error prone handcrafted features. While deep learning has shown success in fields such as image classification and natural language processing, its application for feature extraction on raw network packet data for intrusion detection is largely unexplored. In this paper we modify a Word2Vec approach, used for text processing, and apply it to packet data for automatic feature extraction. We call this approach Packet2Vec. For the classification task of benign versus malicious traffic on a 2009 DARPA network data set, we obtain an area under the curve (AUC) of the receiver operating characteristic (ROC) between 0.988-0.996 and an AUC of the Precision/Recall curve between 0.604-0.667.

arXiv:2004.13799

Minority Reports Defense: Defending Against Adversarial Patches

Deep learning image classification is vulnerable to adversarial attack, even if the attacker changes just a small patch of the image. We propose a defense against patch attacks based on partially occluding the image around each candidate patch location, so that a few occlusions each completely hide the patch. We demonstrate on CIFAR-10, Fashion MNIST, and MNIST that our defense provides certified security against patch attacks of a certain size.

arXiv:2004.14178

Deepfake Video Forensics based on Transfer Learning

Deeplearning has been used to solve complex problems in various domains. As it advances, it also creates applications which become a major threat to our privacy, security and even to our Democracy. Such an application which is being developed recently is the "Deepfake". Deepfake models can create fake images and videos that humans cannot differentiate them from the genuine ones. Therefore, the counter application to automatically detect and analyze the digital visual media is necessary in today world. This paper details retraining the image classification models to apprehend the features from each deepfake video frames. After feeding different sets of deepfake clips of video fringes through a pretrained layer of bottleneck in the neural network is made for every video frame, already stated layer contains condense data for all images and exposes artificial manipulations in Deepfake videos. When checking Deepfake videos, this technique received more than 87 per cent accuracy. This technique has been tested on the Face Forensics dataset and obtained good accuracy in detection.

arXiv:2004.14152공식 코드 · Papers with Code

A Fast 3D CNN for Hyperspectral Image Classification

Hyperspectral imaging (HSI) has been extensively utilized for a number of real-world applications. HSI classification (HSIC) is a challenging task due to high inter-class similarity, high intra-class variability, overlapping, and nested regions. A 2D Convolutional Neural Network (CNN) is a viable approach whereby HSIC highly depends on both Spectral-Spatial information, therefore, 3D CNN can be an alternative but highly computational complex due to the volume and spectral dimensions. Furthermore, these models do not extract quality feature maps and may underperform over the regions having similar textures. Therefore, this work proposed a 3D CNN model that utilizes both spatial-spectral feature maps to attain good performance. In order to achieve the said performance, the HSI cube is first divided into small overlapping 3D patches. Later these patches are processed to generate 3D feature maps using a 3D kernel function over multiple contiguous bands that persevere the spectral information as well. Benchmark HSI datasets (Pavia University, Salinas and Indian Pines) are considered to validate the performance of our proposed method. The results are further compared with several state-of-the-art methods.

arXiv:2004.14368

VGGSound: A Large-scale Audio-Visual Dataset

Our goal is to collect a large-scale audio-visual dataset with low label noise from videos in the wild using computer vision techniques. The resulting dataset can be used for training and evaluating audio recognition models. We make three contributions. First, we propose a scalable pipeline based on computer vision techniques to create an audio dataset from open-source media. Our pipeline involves obtaining videos from YouTube; using image classification algorithms to localize audio-visual correspondence; and filtering out ambient noise using audio verification. Second, we use this pipeline to curate the VGGSound dataset consisting of more than 210k videos for 310 audio classes. Third, we investigate various Convolutional Neural Network~(CNN) architectures and aggregation approaches to establish audio recognition baselines for our new dataset. Compared to existing audio datasets, VGGSound ensures audio-visual correspondence and is collected under unconstrained conditions. Code and the dataset are available at http://www.robots.ox.ac.uk/~vgg/data/vggsound/

arXiv:2004.14340공식 코드 · Papers with Code

WoodFisher: Efficient Second-Order Approximation for Neural Network Compression

Second-order information, in the form of Hessian- or Inverse-Hessian-vector products, is a fundamental tool for solving optimization problems. Recently, there has been significant interest in utilizing this information in the context of deep neural networks; however, relatively little is known about the quality of existing approximations in this context. Our work examines this question, identifies issues with existing approaches, and proposes a method called WoodFisher to compute a faithful and efficient estimate of the inverse Hessian. Our main application is to neural network compression, where we build on the classic Optimal Brain Damage/Surgeon framework. We demonstrate that WoodFisher significantly outperforms popular state-of-the-art methods for one-shot pruning. Further, even when iterative, gradual pruning is considered, our method results in a gain in test accuracy over the state-of-the-art approaches, for pruning popular neural networks (like ResNet-50, MobileNetV1) trained on standard image classification datasets such as ImageNet ILSVRC. We examine how our method can be extended to take into account first-order information, as well as illustrate its ability to automatically set layer-wise pruning thresholds and perform compression in the limited-data regime. The code is available at the following link, https://github.com/IST-DASLab/WoodFisher.

arXiv:2004.13532

Integration of Leaky-Integrate-and-Fire-Neurons in Deep Learning Architectures

Up to now, modern Machine Learning is mainly based on fitting high dimensional functions to enormous data sets, taking advantage of huge hardware resources. We show that biologically inspired neuron models such as the Leaky-Integrate-and-Fire (LIF) neurons provide novel and efficient ways of information encoding. They can be integrated in Machine Learning models, and are a potential target to improve Machine Learning performance. Thus, we derived simple update-rules for the LIF units from the differential equations, which are easy to numerically integrate. We apply a novel approach to train the LIF units supervisedly via backpropagation, by assigning a constant value to the derivative of the neuron activation function exclusively for the backpropagation step. This simple mathematical trick helps to distribute the error between the neurons of the pre-connected layer. We apply our method to the IRIS blossoms image data set and show that the training technique can be used to train LIF neurons on image classification tasks. Furthermore, we show how to integrate our method in the KERAS (tensorflow) framework and efficiently run it on GPUs. To generate a deeper understanding of the mechanisms during training we developed interactive illustrations, which we provide online. With this study we want to contribute to the current efforts to enhance Machine Intelligence by integrating principles from biology.

arXiv:2004.13414

Pseudo Rehearsal using non photo-realistic images

Deep Neural networks forget previously learnt tasks when they are faced with learning new tasks. This is called catastrophic forgetting. Rehearsing the neural network with the training data of the previous task can protect the network from catastrophic forgetting. Since rehearsing requires the storage of entire previous data, Pseudo rehearsal was proposed, where samples belonging to the previous data are generated synthetically for rehearsal. In an image classification setting, while current techniques try to generate synthetic data that is photo-realistic, we demonstrated that Neural networks can be rehearsed on data that is not photo-realistic and still achieve good retention of the previous task. We also demonstrated that forgoing the constraint of having photo realism in the generated data can result in a significant reduction in the consumption of computational and memory resources for pseudo rehearsal.

arXiv:2004.13406

Identification of Cervical Pathology using Adversarial Neural Networks

Various screening and diagnostic methods have led to a large reduction of cervical cancer death rates in developed countries. However, cervical cancer is the leading cause of cancer related deaths in women in India and other low and middle income countries (LMICs) especially among the urban poor and slum dwellers. Several sophisticated techniques such as cytology tests, HPV tests etc. have been widely used for screening of cervical cancer. These tests are inherently time consuming. In this paper, we propose a convolutional autoencoder based framework, having an architecture similar to SegNet which is trained in an adversarial fashion for classifying images of the cervix acquired using a colposcope. We validate performance on the Intel-Mobile ODT cervical image classification dataset. The proposed method outperforms the standard technique of fine-tuning convolutional neural networks pre-trained on ImageNet database with an average accuracy of 73.75%.

arXiv:2004.13271

Trainable Activation Function in Image Classification

In the current research of neural networks, the activation function is manually specified by human and not able to change themselves during training. This paper focus on how to make the activation function trainable for deep neural networks. We use series and linear combination of different activation functions make activation functions continuously variable. Also, we test the performance of CNNs with Fourier series simulated activation(Fourier-CNN) and CNNs with linear combined activation function (LC-CNN) on Cifar-10 dataset. The result shows our trainable activation function reveals better performance than the most used ReLU activation function. Finally, we improves the performance of Fourier-CNN with Autoencoder, and test the performance of PSO algorithm in optimizing the parameters of networks

arXiv:2004.11995

Explicit Domain Adaptation with Loosely Coupled Samples

Transfer learning is an important field of machine learning in general, and particularly in the context of fully autonomous driving, which needs to be solved simultaneously for many different domains, such as changing weather conditions and country-specific driving behaviors. Traditional transfer learning methods often focus on image data and are black-box models. In this work we propose a transfer learning framework, core of which is learning an explicit mapping between domains. Due to its interpretability, this is beneficial for safety-critical applications, like autonomous driving. We show its general applicability by considering image classification problems and then move on to time-series data, particularly predicting lane changes. In our evaluation we adapt a pre-trained model to a dataset exhibiting different driving and sensory characteristics.

arXiv:2004.12385공식 코드 · Papers with Code

Towards Feature Space Adversarial Attack

We propose a new adversarial attack to Deep Neural Networks for image classification. Different from most existing attacks that directly perturb input pixels, our attack focuses on perturbing abstract features, more specifically, features that denote styles, including interpretable styles such as vivid colors and sharp outlines, and uninterpretable ones. It induces model misclassfication by injecting imperceptible style changes through an optimization procedure. We show that our attack can generate adversarial samples that are more natural-looking than the state-of-the-art unbounded attacks. The experiment also supports that existing pixel-space adversarial attack detection and defense techniques can hardly ensure robustness in the style related feature space.

arXiv:2004.12381

Hyperspectral Images Classification Based on Multi-scale Residual Network

Because hyperspectral remote sensing images contain a lot of redundant information and the data structure is highly non-linear, leading to low classification accuracy of traditional machine learning methods. The latest research shows that hyperspectral image classification based on deep convolutional neural network has high accuracy. However, when a small amount of data is used for training, the classification accuracy of deep learning methods is greatly reduced. In order to solve the problem of low classification accuracy of existing algorithms on small samples of hyperspectral images, a multi-scale residual network is proposed. The multi-scale extraction and fusion of spatial and spectral features is realized by adding a branch structure into the residual block and using convolution kernels of different sizes in the branch. The spatial and spectral information contained in hyperspectral images are fully utilized to improve the classification accuracy. In addition, in order to improve the speed and prevent overfitting, the model uses dynamic learning rate, BN and Dropout strategies. The experimental results show that the overall classification accuracy of this method is 99.07% and 99.96% respectively in the data set of Indian Pines and Pavia University, which is better than other algorithms.

arXiv:2004.11805

Applications of the Streaming Networks

Most recently Streaming Networks (STnets) have been introduced as a mechanism of robust noise-corrupted images classification. STnets is a family of convolutional neural networks, which consists of multiple neural networks (streams), which have different inputs and their outputs are concatenated and fed into a single joint classifier. The original paper has illustrated how STnets can successfully classify images from Cifar10, EuroSat and UCmerced datasets, when images were corrupted with various levels of random zero noise. In this paper, we demonstrate that STnets are capable of high accuracy classification of images corrupted with Gaussian noise, fog, snow, etc. (Cifar10 corrupted dataset) and low light images (subset of Carvana dataset). We also introduce a new type of STnets called Hybrid STnets. Thus, we illustrate that STnets is a universal tool of image classification when original training dataset is corrupted with noise or other transformations, which lead to information loss from original images.