PyTorchKR Papers

최신 논문

총 8,922편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2203.10939

Deep Learning Serves Traffic Safety Analysis: A Forward-looking Review

This paper explores Deep Learning (DL) methods that are used or have the potential to be used for traffic video analysis, emphasizing driving safety for both Autonomous Vehicles (AVs) and human-operated vehicles. We present a typical processing pipeline, which can be used to understand and interpret traffic videos by extracting operational safety metrics and providing general hints and guidelines to improve traffic safety. This processing framework includes several steps, including video enhancement, video stabilization, semantic and incident segmentation, object detection and classification, trajectory extraction, speed estimation, event analysis, modeling and anomaly detection. Our main goal is to guide traffic analysts to develop their own custom-built processing frameworks by selecting the best choices for each step and offering new designs for the lacking modules by providing a comparative analysis of the most successful conventional and DL-based algorithms proposed for each step. We also review existing open-source tools and public datasets that can help train DL models. To be more specific, we review exemplary traffic problems and mentioned requires steps for each problem. Besides, we investigate connections to the closely related research areas of drivers' cognition evaluation, Crowd-sourcing-based monitoring systems, Edge Computing in roadside infrastructures, Automated Driving Systems (ADS)-equipped vehicles, and highlight the missing gaps. Finally, we review commercial implementations of traffic monitoring systems, their future outlook, and open problems and remaining challenges for widespread use of such systems.

arXiv:2207.01768

Rank-Based Filter Pruning for Real-Time UAV Tracking

Unmanned aerial vehicle (UAV) tracking has wide potential applications in such as agriculture, navigation, and public security. However, the limitations of computing resources, battery capacity, and maximum load of UAV hinder the deployment of deep learning-based tracking algorithms on UAV. Consequently, discriminative correlation filters (DCF) trackers stand out in the UAV tracking community because of their high efficiency. However, their precision is usually much lower than trackers based on deep learning. Model compression is a promising way to narrow the gap (i.e., effciency, precision) between DCF- and deep learning- based trackers, which has not caught much attention in UAV tracking. In this paper, we propose the P-SiamFC++ tracker, which is the first to use rank-based filter pruning to compress the SiamFC++ model, achieving a remarkable balance between efficiency and precision. Our method is general and may encourage further studies on UAV tracking with model compression. Extensive experiments on four UAV benchmarks, including UAV123@10fps, DTB70, UAVDT and Vistrone2018, show that P-SiamFC++ tracker significantly outperforms state-of-the-art UAV tracking methods.

arXiv:2207.02163

Automatic inspection of cultural monuments using deep and tensor-based learning on hyperspectral imagery

In Cultural Heritage, hyperspectral images are commonly used since they provide extended information regarding the optical properties of materials. Thus, the processing of such high-dimensional data becomes challenging from the perspective of machine learning techniques to be applied. In this paper, we propose a Rank-$R$ tensor-based learning model to identify and classify material defects on Cultural Heritage monuments. In contrast to conventional deep learning approaches, the proposed high order tensor-based learning demonstrates greater accuracy and robustness against overfitting. Experimental results on real-world data from UNESCO protected areas indicate the superiority of the proposed scheme compared to conventional deep learning models.

arXiv:2207.02073

A Densely Interconnected Network for Deep Learning Accelerated MRI

Objective: To improve accelerated MRI reconstruction through a densely connected cascading deep learning reconstruction framework. Materials and Methods: A cascading deep learning reconstruction framework (baseline model) was modified by applying three architectural modifications: Input-level dense connections between cascade inputs and outputs, an improved deep learning sub-network, and long-range skip-connections between subsequent deep learning networks. An ablation study was performed, where five model configurations were trained on the NYU fastMRI neuro dataset with an end-to-end scheme conjunct on four- and eight-fold acceleration. The trained models were evaluated by comparing their respective structural similarity index measure (SSIM), normalized mean square error (NMSE) and peak signal to noise ratio (PSNR). Results: The proposed densely interconnected residual cascading network (DIRCN), utilizing all three suggested modifications, achieved a SSIM improvement of 8% and 11% for four- and eight-fold acceleration, respectively. For eight-fold acceleration, the model achieved a 23% decrease in the NMSE when compared to the baseline model. In an ablation study, the individual architectural modifications all contributed to this improvement, by reducing the SSIM and NMSE with approximately 3% and 5% for four-fold acceleration, respectively. Conclusion: The proposed architectural modifications allow for simple adjustments on an already existing cascading framework to further improve the resulting reconstructions.

arXiv:2207.02000

Disentangling private classes through regularization

Deep learning models are nowadays broadly deployed to solve an incredibly large variety of tasks. However, little attention has been devoted to connected legal aspects. In 2016, the European Union approved the General Data Protection Regulation which entered into force in 2018. Its main rationale was to protect the privacy and data protection of its citizens by the way of operating of the so-called "Data Economy". As data is the fuel of modern Artificial Intelligence, it is argued that the GDPR can be partly applicable to a series of algorithmic decision making tasks before a more structured AI Regulation enters into force. In the meantime, AI should not allow undesired information leakage deviating from the purpose for which is created. In this work we propose DisP, an approach for deep learning models disentangling the information related to some classes we desire to keep private, from the data processed by AI. In particular, DisP is a regularization strategy de-correlating the features belonging to the same private class at training time, hiding the information of private classes membership. Our experiments on state-of-the-art deep learning models show the effectiveness of DisP, minimizing the risk of extraction for the classes we desire to keep private.

arXiv:2207.01811

Deriving Surface Resistivity from Polarimetric SAR Data Using Dual-Input UNet

Traditional survey methods for finding surface resistivity are time-consuming and labor intensive. Very few studies have focused on finding the resistivity/conductivity using remote sensing data and deep learning techniques. In this line of work, we assessed the correlation between surface resistivity and Synthetic Aperture Radar (SAR) by applying various deep learning methods and tested our hypothesis in the Coso Geothermal Area, USA. For detecting the resistivity, L-band full polarimetric SAR data acquired by UAVSAR were used, and MT (Magnetotellurics) inverted resistivity data of the area were used as the ground truth. We conducted experiments to compare various deep learning architectures and suggest the use of Dual Input UNet (DI-UNet) architecture. DI-UNet uses a deep learning architecture to predict the resistivity using full polarimetric SAR data by promising a quick survey addition to the traditional method. Our proposed approach accomplished improved outcomes for the mapping of MT resistivity from SAR data.

arXiv:2207.00975

Understanding Tieq Viet with Deep Learning Models

Deep learning is a powerful approach in recovering lost information as well as harder inverse function computation problems. When applied in natural language processing, this approach is essentially making use of context as a mean to recover information through likelihood maximization. Not long ago, a linguistic study called Tieq Viet was controversial among both researchers and society. We find this a great example to demonstrate the ability of deep learning models to recover lost information. In the proposal of Tieq Viet, some consonants in the standard Vietnamese are replaced. A sentence written in this proposal can be interpreted into multiple sentences in the standard version, with different meanings. The hypothesis that we want to test is whether a deep learning model can recover the lost information if we translate the text from Vietnamese to Tieq Viet.

arXiv:2207.00808

On the modern deep learning approaches for precipitation downscaling

Deep Learning (DL) based downscaling has become a popular tool in earth sciences recently. Increasingly, different DL approaches are being adopted to downscale coarser precipitation data and generate more accurate and reliable estimates at local (~few km or even smaller) scales. Despite several studies adopting dynamical or statistical downscaling of precipitation, the accuracy is limited by the availability of ground truth. A key challenge to gauge the accuracy of such methods is to compare the downscaled data to point-scale observations which are often unavailable at such small scales. In this work, we carry out the DL-based downscaling to estimate the local precipitation data from the India Meteorological Department (IMD), which was created by approximating the value from station location to a grid point. To test the efficacy of different DL approaches, we apply four different methods of downscaling and evaluate their performance. The considered approaches are (i) Deep Statistical Downscaling (DeepSD), augmented Convolutional Long Short Term Memory (ConvLSTM), fully convolutional network (U-NET), and Super-Resolution Generative Adversarial Network (SR-GAN). A custom VGG network, used in the SR-GAN, is developed in this work using precipitation data. The results indicate that SR-GAN is the best method for precipitation data downscaling. The downscaled data is validated with precipitation values at IMD station. This DL method offers a promising alternative to statistical downscaling.

arXiv:2207.00739

Deep Learning for Systemic Risk Measures

The aim of this paper is to study a new methodological framework for systemic risk measures by applying deep learning method as a tool to compute the optimal strategy of capital allocations. Under this new framework, systemic risk measures can be interpreted as the minimal amount of cash that secures the aggregated system by allocating capital to the single institutions before aggregating the individual risks. This problem has no explicit solution except in very limited situations. Deep learning is increasingly receiving attention in financial modelings and risk management and we propose our deep learning based algorithms to solve both the primal and dual problems of the risk measures, and thus to learn the fair risk allocations. In particular, our method for the dual problem involves the training philosophy inspired by the well-known Generative Adversarial Networks (GAN) approach and a newly designed direct estimation of Radon-Nikodym derivative. We close the paper with substantial numerical studies of the subject and provide interpretations of the risk allocations associated to the systemic risk measures. In the particular case of exponential preferences, numerical experiments demonstrate excellent performance of the proposed algorithm, when compared with the optimal explicit solution as a benchmark.

arXiv:2207.01581

Interpretable Fusion Analytics Framework for fMRI Connectivity: Self-Attention Mechanism and Latent Space Item-Response Model

There have been several attempts to use deep learning based on brain fMRI signals to classify cognitive impairment diseases. However, deep learning is a hidden black box model that makes it difficult to interpret the process of classification. To address this issue, we propose a novel analytical framework that interprets the classification result from deep learning processes. We first derive the region of interest (ROI) functional connectivity network (FCN) by embedding functions based on their similar signal patterns. Then, using the self-attention equipped deep learning model, we classify diseases based on their FCN. Finally, in order to interpret the classification results, we employ a latent space item-response interaction network model to identify the significant functions that exhibit distinct connectivity patterns when compared to other diseases. The application of this proposed framework to the four types of cognitive impairment shows that our approach is valid for determining the significant ROI functions.

arXiv:2207.00529공식 코드 · Papers with Code

Deep Learning and Symbolic Regression for Discovering Parametric Equations

Symbolic regression is a machine learning technique that can learn the governing formulas of data and thus has the potential to transform scientific discovery. However, symbolic regression is still limited in the complexity and dimensionality of the systems that it can analyze. Deep learning on the other hand has transformed machine learning in its ability to analyze extremely complex and high-dimensional datasets. We propose a neural network architecture to extend symbolic regression to parametric systems where some coefficient may vary but the structure of the underlying governing equation remains constant. We demonstrate our method on various analytic expressions, ODEs, and PDEs with varying coefficients and show that it extrapolates well outside of the training domain. The neural network-based architecture can also integrate with other deep learning architectures so that it can analyze high-dimensional data while being trained end-to-end. To this end we integrate our architecture with convolutional neural networks to analyze 1D images of varying spring systems.

arXiv:2207.00524

Solving barrier options under stochastic volatility using deep learning

We develop an unsupervised deep learning method to solve the barrier options under the Bergomi model. The neural networks serve as the approximate option surfaces and are trained to satisfy the PDE as well as the boundary conditions. Two singular terms are added to the neural networks to deal with the non-smooth and discontinuous payoff at the strike and barrier levels so that the neural networks can replicate the asymptotic behaviors of barrier options at short maturities. After that, vanilla options and barrier options are priced in a single framework. Also, neural networks are employed to deal with the high dimensionality of the function input in the Bergomi model. Once trained, the neural network solution yields fast and accurate option values.

arXiv:2207.00336

A Deep-Learning-Aided Pipeline for Efficient Post-Silicon Tuning

In post-silicon validation, tuning is to find the values for the tuning knobs, potentially as a function of process parameters and/or known operating conditions. In this sense, an more efficient tuning requires identifying the most critical tuning knobs and process parameters in terms of a given figure-of-merit for a Device Under Test (DUT). This is often manually conducted by experienced experts. However, with increasingly complex chips, manual inspection on a large amount of raw variables has become more challenging. In this work, we leverage neural networks to efficiently select the most relevant variables and present a corresponding deep-learning-aided pipeline for efficient tuning.

arXiv:2207.00237공식 코드 · Papers with Code

Improving Speech Enhancement through Fine-Grained Speech Characteristics

While deep learning based speech enhancement systems have made rapid progress in improving the quality of speech signals, they can still produce outputs that contain artifacts and can sound unnatural. We propose a novel approach to speech enhancement aimed at improving perceptual quality and naturalness of enhanced signals by optimizing for key characteristics of speech. We first identify key acoustic parameters that have been found to correlate well with voice quality (e.g. jitter, shimmer, and spectral flux) and then propose objective functions which are aimed at reducing the difference between clean speech and enhanced speech with respect to these features. The full set of acoustic features is the extended Geneva Acoustic Parameter Set (eGeMAPS), which includes 25 different attributes associated with perception of speech. Given the non-differentiable nature of these feature computation, we first build differentiable estimators of the eGeMAPS and then use them to fine-tune existing speech enhancement systems. Our approach is generic and can be applied to any existing deep learning based enhancement systems to further improve the enhanced speech signals. Experimental results conducted on the Deep Noise Suppression (DNS) Challenge dataset shows that our approach can improve the state-of-the-art deep learning based enhancement systems.

arXiv:2207.00137

Robustness of Epinets against Distributional Shifts

Recent work introduced the epinet as a new approach to uncertainty modeling in deep learning. An epinet is a small neural network added to traditional neural networks, which, together, can produce predictive distributions. In particular, using an epinet can greatly improve the quality of joint predictions across multiple inputs, a measure of how well a neural network knows what it does not know. In this paper, we examine whether epinets can offer similar advantages under distributional shifts. We find that, across ImageNet-A/O/C, epinets generally improve robustness metrics. Moreover, these improvements are more significant than those afforded by even very large ensembles at orders of magnitude lower computational costs. However, these improvements are relatively small compared to the outstanding issues in distributionally-robust deep learning. Epinets may be a useful tool in the toolbox, but they are far from the complete solution.

arXiv:2206.15192

Privacy-preserving household load forecasting based on non-intrusive load monitoring: A federated deep learning approach

Load forecasting is very essential in the analysis and grid planning of power systems. For this reason, we first propose a household load forecasting method based on federated deep learning and non-intrusive load monitoring (NILM). For all we know, this is the first research on federated learning (FL) in household load forecasting based on NILM. In this method, the integrated power is decomposed into individual device power by non-intrusive load monitoring, and the power of individual appliances is predicted separately using a federated deep learning model. Finally, the predicted power values of individual appliances are aggregated to form the total power prediction. Specifically, by separately predicting the electrical equipment to obtain the predicted power, it avoids the error caused by the strong time dependence in the power signal of a single device. And in the federated deep learning prediction model, the household owners with the power data share the parameters of the local model instead of the local power data, guaranteeing the privacy of the household user data. The case results demonstrate that the proposed approach provides a better prediction effect than the traditional methodology that directly predicts the aggregated signal as a whole. In addition, experiments in various federated learning environments are designed and implemented to validate the validity of this methodology.

arXiv:2206.15030

Modern Question Answering Datasets and Benchmarks: A Survey

Question Answering (QA) is one of the most important natural language processing (NLP) tasks. It aims using NLP technologies to generate a corresponding answer to a given question based on the massive unstructured corpus. With the development of deep learning, more and more challenging QA datasets are being proposed, and lots of new methods for solving them are also emerging. In this paper, we investigate influential QA datasets that have been released in the era of deep learning. Specifically, we begin with introducing two of the most common QA tasks - textual question answer and visual question answering - separately, covering the most representative datasets, and then give some current challenges of QA research.

arXiv:2206.15351

Deep Learning to See: Towards New Foundations of Computer Vision

The remarkable progress in computer vision over the last few years is, by and large, attributed to deep learning, fueled by the availability of huge sets of labeled data, and paired with the explosive growth of the GPU paradigm. While subscribing to this view, this book criticizes the supposed scientific progress in the field and proposes the investigation of vision within the framework of information-based laws of nature. Specifically, the present work poses fundamental questions about vision that remain far from understood, leading the reader on a journey populated by novel challenges resonating with the foundations of machine learning. The central thesis is that for a deeper understanding of visual computational processes, it is necessary to look beyond the applications of general purpose machine learning algorithms and focus instead on appropriate learning theories that take into account the spatiotemporal nature of the visual signal.

arXiv:2206.14373

Theoretical Perspectives on Deep Learning Methods in Inverse Problems

In recent years, there have been significant advances in the use of deep learning methods in inverse problems such as denoising, compressive sensing, inpainting, and super-resolution. While this line of works has predominantly been driven by practical algorithms and experiments, it has also given rise to a variety of intriguing theoretical problems. In this paper, we survey some of the prominent theoretical developments in this line of works, focusing in particular on generative priors, untrained neural network priors, and unfolding algorithms. In addition to summarizing existing results in these topics, we highlight several ongoing challenges and open problems.

arXiv:2206.14322공식 코드 · Papers with Code

An Empirical Study of Challenges in Converting Deep Learning Models

There is an increase in deploying Deep Learning (DL)-based software systems in real-world applications. Usually DL models are developed and trained using DL frameworks that have their own internal mechanisms/formats to represent and train DL models, and usually those formats cannot be recognized by other frameworks. Moreover, trained models are usually deployed in environments different from where they were developed. To solve the interoperability issue and make DL models compatible with different frameworks/environments, some exchange formats are introduced for DL models, like ONNX and CoreML. However, ONNX and CoreML were never empirically evaluated by the community to reveal their prediction accuracy, performance, and robustness after conversion. Poor accuracy or non-robust behavior of converted models may lead to poor quality of deployed DL-based software systems. We conduct, in this paper, the first empirical study to assess ONNX and CoreML for converting trained DL models. In our systematic approach, two popular DL frameworks, Keras and PyTorch, are used to train five widely used DL models on three popular datasets. The trained models are then converted to ONNX and CoreML and transferred to two runtime environments designated for such formats, to be evaluated. We investigate the prediction accuracy before and after conversion. Our results unveil that the prediction accuracy of converted models are at the same level of originals. The performance (time cost and memory consumption) of converted models are studied as well. The size of models are reduced after conversion, which can result in optimized DL-based software deployment. Converted models are generally assessed as robust at the same level of originals. However, obtained results show that CoreML models are more vulnerable to adversarial attacks compared to ONNX.