PyTorchKR Papers

최신 논문

총 8,922편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

논문 ID: document/9274427

Effective Deep Learning Models for Automatic Diacritization of Arabic Text

While building a text-to-speech system for the Arabic language, we found that the system synthesized speeches with many pronunciation errors. The primary source of these errors is the lack of diacritics in modern standard Arabic writing. These diacritics are small strokes that appear above or below each letter to provide pronunciation and grammatical information. We propose three deep learning models to recover Arabic text diacritics based on our work in a text-to-speech synthesis system using deep learning. The first model is a baseline model used to test how a simple deep learning model performs on the corpora. The second model is based on an encoder-decoder architecture, which resembles our text-to-speech synthesis model with many modifications to suit this problem. The last model is based on the encoder part of the text-to-speech model, which achieves state-of-the-art performances in both word error rate and diacritic error rate metrics. These models will benefit a wide range of natural language processing applications such as text-to-speech, part-of-speech tagging, and machine translation.

arXiv:2103.09316공식 코드 · Papers with Code

Are deep learning models superior for missing data imputation in large surveys? Evidence from an empirical comparison

Multiple imputation (MI) is a popular approach for dealing with missing data arising from non-response in sample surveys. Multiple imputation by chained equations (MICE) is one of the most widely used MI algorithms for multivariate data, but it lacks theoretical foundation and is computationally intensive. Recently, missing data imputation methods based on deep learning models have been developed with encouraging results in small studies. However, there has been limited research on evaluating their performance in realistic settings compared to MICE, particularly in big surveys. We conduct extensive simulation studies based on a subsample of the American Community Survey to compare the repeated sampling properties of four machine learning based MI methods: MICE with classification trees, MICE with random forests, generative adversarial imputation networks, and multiple imputation using denoising autoencoders. We find the deep learning imputation methods are superior to MICE in terms of computational time. However, with the default choice of hyperparameters in the common software packages, MICE with classification trees consistently outperforms, often by a large margin, the deep learning imputation methods in terms of bias, mean squared error, and coverage under a range of realistic settings.

arXiv:2103.09743

Deep Learning-based Extreme Heatwave Forecast

Because of the impact of extreme heat waves and heat domes on society and biodiversity, their study is a key challenge. We specifically study long-lasting extreme heat waves, which are among the most important for climate impacts. Physics driven weather forecast systems or climate models can be used to forecast their occurrence or predict their probability. The present work explores the use of deep learning architectures, trained using outputs of a climate model, as an alternative strategy to forecast the occurrence of extreme long-lasting heatwaves. This new approach will be useful for several key scientific goals which include the study of climate model statistics, building a quantitative proxy for resampling rare events in climate models, study the impact of climate change, and should eventually be useful for forecasting. Fulfilling these important goals implies addressing issues such as class-size imbalance that is intrinsically associated with rare event prediction, assessing the potential benefits of transfer learning to address the nested nature of extreme events (naturally included in less extreme ones). We train a Convolutional Neural Network, using 1000 years of climate model outputs, with large-class undersampling and transfer learning. From the observed snapshots of the surface temperature and the 500 hPa geopotential height fields, the trained network achieves significant performance in forecasting the occurrence of long-lasting extreme heatwaves. We are able to predict them at three different levels of intensity, and as early as 15 days ahead of the start of the event (30 days ahead of the end of the event).

arXiv:2103.09708공식 코드 · Papers with Code

What's in My LiDAR Odometry Toolbox?

With the democratization of 3D LiDAR sensors, precise LiDAR odometries and SLAM are in high demand. New methods regularly appear, proposing solutions ranging from small variations in classical algorithms to radically new paradigms based on deep learning. Yet it is often difficult to compare these methods, notably due to the few datasets on which the methods can be evaluated and compared. Furthermore, their weaknesses are rarely examined, often letting the user discover the hard way whether a method would be appropriate for a use case. In this paper, we review and organize the main 3D LiDAR odometries into distinct categories. We implemented several approaches (geometric based, deep learning based, and hybrid methods) to conduct an in-depth analysis of their strengths and weaknesses on multiple datasets, guiding the reader through the different LiDAR odometries available. Implementation of the methods has been made publicly available at https://github.com/Kitware/pyLiDAR-SLAM.

arXiv:2103.09300

The impact of data volume on performance of deep learning based building rooftop extraction using very high spatial resolution aerial images

Building rooftop data are of importance in several urban applications and in natural disaster management. In contrast to traditional surveying and mapping, by using high spatial resolution aerial images, deep learning-based building rooftops extraction methods are efficient and accurate. Although more training data is preferred in deep learning-based tasks, the effect of data volume on building extraction models is underexplored. Therefore, the paper explores the impact of data volume on the performance of building rooftop extraction from very-high-spatial-resolution (VHSR) images using deep learning-based methods. To do so, we manually labelled 0.12m spatial resolution aerial images and perform a comparative analysis of models trained on datasets of different sizes using popular deep learning architectures for segmentation tasks, including Fully Convolutional Networks (FCN)-8s, U-Net and DeepLabv3+. The experiments showed that with more training data, algorithms converged faster and achieved higher accuracy, while better algorithms were able to better mitigate the lack of training data.

arXiv:2103.08894

Distributed Deep Learning Using Volunteer Computing-Like Paradigm

Use of Deep Learning (DL) in commercial applications such as image classification, sentiment analysis and speech recognition is increasing. When training DL models with large number of parameters and/or large datasets, cost and speed of training can become prohibitive. Distributed DL training solutions that split a training job into subtasks and execute them over multiple nodes can decrease training time. However, the cost of current solutions, built predominantly for cluster computing systems, can still be an issue. In contrast to cluster computing systems, Volunteer Computing (VC) systems can lower the cost of computing, but applications running on VC systems have to handle fault tolerance, variable network latency and heterogeneity of compute nodes, and the current solutions are not designed to do so. We design a distributed solution that can run DL training on a VC system by using a data parallel approach. We implement a novel asynchronous SGD scheme called VC-ASGD suited for VC systems. In contrast to traditional VC systems that lower cost by using untrustworthy volunteer devices, we lower cost by leveraging preemptible computing instances on commercial cloud platforms. By using preemptible instances that require applications to be fault tolerant, we lower cost by 70-90% and improve data security.

arXiv:2103.08870

Learned Gradient Compression for Distributed Deep Learning

Training deep neural networks on large datasets containing high-dimensional data requires a large amount of computation. A solution to this problem is data-parallel distributed training, where a model is replicated into several computational nodes that have access to different chunks of the data. This approach, however, entails high communication rates and latency because of the computed gradients that need to be shared among nodes at every iteration. The problem becomes more pronounced in the case that there is wireless communication between the nodes (i.e. due to the limited network bandwidth). To address this problem, various compression methods have been proposed including sparsification, quantization, and entropy encoding of the gradients. Existing methods leverage the intra-node information redundancy, that is, they compress gradients at each node independently. In contrast, we advocate that the gradients across the nodes are correlated and propose methods to leverage this inter-node redundancy to improve compression efficiency. Depending on the node communication protocol (parameter server or ring-allreduce), we propose two instances of the LGC approach that we coin Learned Gradient Compression (LGC). Our methods exploit an autoencoder (i.e. trained during the first stages of the distributed training) to capture the common information that exists in the gradients of the distributed nodes. We have tested our LGC methods on the image classification and semantic segmentation tasks using different convolutional neural networks (ResNet50, ResNet101, PSPNet) and multiple datasets (ImageNet, Cifar10, CamVid). The ResNet101 model trained for image classification on Cifar10 achieved an accuracy of 93.57%, which is lower than the baseline distributed training with uncompressed gradients only by 0.18%.

arXiv:2103.08700

Deep Learning for Chest X-ray Analysis: A Survey

Recent advances in deep learning have led to a promising performance in many medical image analysis tasks. As the most commonly performed radiological exam, chest radiographs are a particularly important modality for which a variety of applications have been researched. The release of multiple, large, publicly available chest X-ray datasets in recent years has encouraged research interest and boosted the number of publications. In this paper, we review all studies using deep learning on chest radiographs, categorizing works by task: image-level prediction (classification and regression), segmentation, localization, image generation and domain adaptation. Commercially available applications are detailed, and a comprehensive discussion of the current state of the art and potential future directions are provided.

arXiv:2103.09154

Leveraging Recent Advances in Deep Learning for Audio-Visual Emotion Recognition

Emotional expressions are the behaviors that communicate our emotional state or attitude to others. They are expressed through verbal and non-verbal communication. Complex human behavior can be understood by studying physical features from multiple modalities; mainly facial, vocal and physical gestures. Recently, spontaneous multi-modal emotion recognition has been extensively studied for human behavior analysis. In this paper, we propose a new deep learning-based approach for audio-visual emotion recognition. Our approach leverages recent advances in deep learning like knowledge distillation and high-performing deep architectures. The deep feature representations of the audio and visual modalities are fused based on a model-level fusion strategy. A recurrent neural network is then used to capture the temporal dynamics. Our proposed approach substantially outperforms state-of-the-art approaches in predicting valence on the RECOLA dataset. Moreover, our proposed visual facial expression feature extraction network outperforms state-of-the-art results on the AffectNet and Google Facial Expression Comparison datasets.

arXiv:2103.09177

Deep learning: a statistical viewpoint

The remarkable practical success of deep learning has revealed some major surprises from a theoretical perspective. In particular, simple gradient methods easily find near-optimal solutions to non-convex optimization problems, and despite giving a near-perfect fit to training data without any explicit effort to control model complexity, these methods exhibit excellent predictive accuracy. We conjecture that specific principles underlie these phenomena: that overparametrization allows gradient methods to find interpolating solutions, that these methods implicitly impose regularization, and that overparametrization leads to benign overfitting. We survey recent theoretical progress that provides examples illustrating these principles in simpler settings. We first review classical uniform convergence results and why they fall short of explaining aspects of the behavior of deep learning methods. We give examples of implicit regularization in simple settings, where gradient methods lead to minimal norm functions that perfectly fit the training data. Then we review prediction methods that exhibit benign overfitting, focusing on regression problems with quadratic loss. For these methods, we can decompose the prediction rule into a simple component that is useful for prediction and a spiky component that is useful for overfitting but, in a favorable setting, does not harm prediction accuracy. We focus specifically on the linear regime for neural networks, where the network can be approximated by a linear model. In this regime, we demonstrate the success of gradient flow, and we consider benign overfitting with two-layer networks, giving an exact asymptotic analysis that precisely demonstrates the impact of overparametrization. We conclude by highlighting the key challenges that arise in extending these insights to realistic deep learning settings.

arXiv:2103.07992

Deep Tiling: Texture Tile Synthesis Using a Deep Learning Approach

Texturing is a fundamental process in computer graphics. Texture is leveraged to enhance the visualization outcome for a 3D scene. In many cases a texture image cannot cover a large 3D model surface because of its small resolution. Conventional techniques like repeating, mirror repeating or clamp to edge do not yield visually acceptable results. Deep learning based texture synthesis has proven to be very effective in such cases. All deep texture synthesis methods trying to create larger resolution textures are limited in terms of GPU memory resources. In this paper, we propose a novel approach to example-based texture synthesis by using a robust deep learning process for creating tiles of arbitrary resolutions that resemble the structural components of an input texture. In this manner, our method is firstly much less memory limited owing to the fact that a new texture tile of small size is synthesized and merged with the original texture and secondly can easily produce missing parts of a large texture.

arXiv:2103.07672공식 코드 · Papers with Code

Fine-grained MRI Reconstruction using Attentive Selection Generative Adversarial Networks

Compressed sensing (CS) leverages the sparsity prior to provide the foundation for fast magnetic resonance imaging (fastMRI). However, iterative solvers for ill-posed problems hinder their adaption to time-critical applications. Moreover, such a prior can be neither rich to capture complicated anatomical structures nor applicable to meet the demand of high-fidelity reconstructions in modern MRI. Inspired by the state-of-the-art methods in image generation, we propose a novel attention-based deep learning framework to provide high-quality MRI reconstruction. We incorporate large-field contextual feature integration and attention selection in a generative adversarial network (GAN) framework. We demonstrate that the proposed model can produce superior results compared to other deep learning-based methods in terms of image quality, and relevance to the MRI reconstruction in an extremely low sampling rate diet.

arXiv:2103.08450

Modeling Multivariate Cyber Risks: Deep Learning Dating Extreme Value Theory

Modeling cyber risks has been an important but challenging task in the domain of cyber security. It is mainly because of the high dimensionality and heavy tails of risk patterns. Those obstacles have hindered the development of statistical modeling of the multivariate cyber risks. In this work, we propose a novel approach for modeling the multivariate cyber risks which relies on the deep learning and extreme value theory. The proposed model not only enjoys the high accurate point predictions via deep learning but also can provide the satisfactory high quantile prediction via extreme value theory. The simulation study shows that the proposed model can model the multivariate cyber risks very well and provide satisfactory prediction performances. The empirical evidence based on real honeypot attack data also shows that the proposed model has very satisfactory prediction performances.

arXiv:2103.08348

Decorrelating Adversarial Nets for Clustering Mobile Network Data

Deep learning will play a crucial role in enabling cognitive automation for the mobile networks of the future. Deep clustering, a subset of deep learning, could be a valuable tool for many network automation use-cases. Unfortunately, most state-of-the-art clustering algorithms target image datasets, which makes them hard to apply to mobile network data due to their highly tuned nature and related assumptions about the data. In this paper, we propose a new algorithm, DANCE (Decorrelating Adversarial Nets for Clustering-friendly Encoding), intended to be a reliable deep clustering method which also performs well when applied to network automation use-cases. DANCE uses a reconstructive clustering approach, separating clustering-relevant from clustering-irrelevant features in a latent representation. This separation removes unnecessary information from the clustering, increasing consistency and peak performance. We comprehensively evaluate DANCE and other select state-of-the-art deep clustering algorithms, and show that DANCE outperforms these algorithms by a significant margin on a mobile network dataset.

arXiv:2103.08259공식 코드 · Papers with Code

The QXS-SAROPT Dataset for Deep Learning in SAR-Optical Data Fusion

Deep learning techniques have made an increasing impact on the field of remote sensing. However, deep neural networks based fusion of multimodal data from different remote sensors with heterogenous characteristics has not been fully explored, due to the lack of availability of big amounts of perfectly aligned multi-sensor image data with diverse scenes of high resolutions, especially for synthetic aperture radar (SAR) data and optical imagery. To promote the development of deep learning based SAR-optical fusion approaches, we release the QXS-SAROPT dataset, which contains 20,000 pairs of SAR-optical image patches. We obtain the SAR patches from SAR satellite GaoFen-3 images and the optical patches from Google Earth images. These images cover three port cities: San Diego, Shanghai and Qingdao. Here, we present a detailed introduction of the construction of the dataset, and show its two representative exemplary applications, namely SAR-optical image matching and SAR ship detection boosted by cross-modal information from optical images. As a large open SAR-optical dataset with multiple scenes of a high resolution, we believe QXS-SAROPT will be of potential value for further research in SAR-optical data fusion technology based on deep learning.

arXiv:2103.07622

Early Prediction and Diagnosis of Retinoblastoma Using Deep Learning Techniques

Retinoblastoma is the most prominent childhood primary intraocular malignancy that impacts the vision of children and adults worldwide. In contrasting and comparing with adults it is uveal melanoma. It is an aggressive tumor that can fill and destroy the eye and the surrounding structures. Therefore early detection of retinoblastoma in childhood is the key. The major impact of the research is to identify the tumor cells in the retina. Also is to find out the stages of the tumor and its corresponding group. The proposed systems assist the ophthalmologists for accurate prediction and diagnosis of retinoblastoma cancer disease at the earliest. The contribution of the proposed approach is to save the life of infants and the grown-up children from vision impairment. The proposed methodology consists of three phases namely, preprocessing, segmentation, and classification. Initially, the fundus images are preprocessed using the Liner Predictive Decision based Median Filter (LPDMF). It removes the noise introduced in the image due to illumination while capturing or scanning the eye of the patients. The preprocessed images are segmented using the Convolutional Neural Network (CNN) to distinguish the foreground tumor cells from the background.

arXiv:2103.07974

CrossoverScheduler: Overlapping Multiple Distributed Training Applications in a Crossover Manner

Distributed deep learning workloads include throughput-intensive training tasks on the GPU clusters, where the Distributed Stochastic Gradient Descent (SGD) incurs significant communication delays after backward propagation, forces workers to wait for the gradient synchronization via a centralized parameter server or directly in decentralized workers. We present CrossoverScheduler, an algorithm that enables communication cycles of a distributed training application to be filled by other applications through pipelining communication and computation. With CrossoverScheduler, the running performance of distributed training can be significantly improved without sacrificing convergence rate and network accuracy. We achieve so by introducing Crossover Synchronization which allows multiple distributed deep learning applications to time-share the same GPU alternately. The prototype of CrossoverScheduler is built and integrated with Horovod. Experiments on a variety of distributed tasks show that CrossoverScheduler achieves 20% \times speedup for image classification tasks on ImageNet dataset.

arXiv:2103.07633공식 코드 · Papers with Code

Attack as Defense: Characterizing Adversarial Examples using Robustness

As a new programming paradigm, deep learning has expanded its application to many real-world problems. At the same time, deep learning based software are found to be vulnerable to adversarial attacks. Though various defense mechanisms have been proposed to improve robustness of deep learning software, many of them are ineffective against adaptive attacks. In this work, we propose a novel characterization to distinguish adversarial examples from benign ones based on the observation that adversarial examples are significantly less robust than benign ones. As existing robustness measurement does not scale to large networks, we propose a novel defense framework, named attack as defense (A2D), to detect adversarial examples by effectively evaluating an example's robustness. A2D uses the cost of attacking an input for robustness evaluation and identifies those less robust examples as adversarial since less robust examples are easier to attack. Extensive experiment results on MNIST, CIFAR10 and ImageNet show that A2D is more effective than recent promising approaches. We also evaluate our defence against potential adaptive attacks and show that A2D is effective in defending carefully designed adaptive attacks, e.g., the attack success rate drops to 0% on CIFAR10.

arXiv:2103.07394공식 코드 · Papers with Code

Despeckling Polarimetric SAR Data Using a Multi-Stream Complex-Valued Fully Convolutional Network

A Polarimetric Synthetic Aperture Radar (PolSAR) sensor is able to collect images in different polarization states, making it a rich source of information for target characterization. PolSAR images are inherently affected by speckle. Therefore, before deriving ad hoc products from the data, the polarimetric covariance matrix needs to be estimated by reducing speckle. In recent years, deep learning based despeckling methods have started to evolve from single channel SAR images to PolSAR images. To this aim, deep learning based approaches separate the real and imaginary components of the complex-valued covariance matrix and use them as independent channels in a standard convolutional neural networks. However, this approach neglects the mathematical relationship that exists between the real and imaginary components, resulting in sub-optimal output. Here, we propose a multi-stream complex-valued fully convolutional network to reduce speckle and effectively estimate the PolSAR covariance matrix. To evaluate the performance of CV-deSpeckNet, we used Sentinel-1 dual polarimetric SAR images to compare against its real-valued counterpart, that separates the real and imaginary parts of the complex covariance matrix. CV-deSpeckNet was also compared against the state of the art PolSAR despeckling methods. The results show CV-deSpeckNet was able to be trained with a fewer number of samples, has a higher generalization capability and resulted in a higher accuracy than its real-valued counterpart and state-of-the-art PolSAR despeckling methods. These results showcase the potential of complex-valued deep learning for PolSAR despeckling.

arXiv:2103.07276

Modelling Animal Biodiversity Using Acoustic Monitoring and Deep Learning

For centuries researchers have used sound to monitor and study wildlife. Traditionally, conservationists have identified species by ear; however, it is now common to deploy audio recording technology to monitor animal and ecosystem sounds. Animals use sound for communication, mating, navigation and territorial defence. Animal sounds provide valuable information and help conservationists to quantify biodiversity. Acoustic monitoring has grown in popularity due to the availability of diverse sensor types which include camera traps, portable acoustic sensors, passive acoustic sensors, and even smartphones. Passive acoustic sensors are easy to deploy and can be left running for long durations to provide insights on habitat and the sounds made by animals and illegal activity. While this technology brings enormous benefits, the amount of data that is generated makes processing a time-consuming process for conservationists. Consequently, there is interest among conservationists to automatically process acoustic data to help speed up biodiversity assessments. Processing these large data sources and extracting relevant sounds from background noise introduces significant challenges. In this paper we outline an approach for achieving this using state of the art in machine learning to automatically extract features from time-series audio signals and modelling deep learning models to classify different bird species based on the sounds they make. The acquired bird songs are processed using mel-frequency cepstrum (MFC) to extract features which are later classified using a multilayer perceptron (MLP). Our proposed method achieved promising results with 0.74 sensitivity, 0.92 specificity and an accuracy of 0.74.