PyTorchKR Papers

최신 논문

총 8,922편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2007.00095공식 코드 · Papers with Code

Deep Learning for Vision-based Prediction: A Survey

Vision-based prediction algorithms have a wide range of applications including autonomous driving, surveillance, human-robot interaction, weather prediction. The objective of this paper is to provide an overview of the field in the past five years with a particular focus on deep learning approaches. For this purpose, we categorize these algorithms into video prediction, action prediction, trajectory prediction, body motion prediction, and other prediction applications. For each category, we highlight the common architectures, training methods and types of data used. In addition, we discuss the common evaluation metrics and datasets used for vision-based prediction tasks. A database of all the information presented in this survey including, cross-referenced according to papers, datasets and metrics, can be found online at https://github.com/aras62/vision-based-prediction.

arXiv:2007.00038공식 코드 · Papers with Code

Unsupervised Deep Learning for Massive MIMO Hybrid Beamforming

Hybrid beamforming is a promising technique to reduce the complexity and cost of massive multiple-input multiple-output (MIMO) systems while providing high data rate. However, the hybrid precoder design is a challenging task requiring channel state information (CSI) feedback and solving a complex optimization problem. This paper proposes a novel RSSI-based unsupervised deep learning method to design the hybrid beamforming in massive MIMO systems. Furthermore, we propose i) a method to design the synchronization signal (SS) in initial access (IA); and ii) a method to design the codebook for the analog precoder. We also evaluate the system performance through a realistic channel model in various scenarios. We show that the proposed method not only greatly increases the spectral efficiency especially in frequency-division duplex (FDD) communication by using partial CSI feedback, but also has near-optimal sum-rate and outperforms other state-of-the-art full-CSI solutions.

arXiv:2007.00523

Drug discovery with explainable artificial intelligence

Deep learning bears promise for drug discovery, including advanced image analysis, prediction of molecular structure and function, and automated generation of innovative chemical entities with bespoke properties. Despite the growing number of successful prospective applications, the underlying mathematical models often remain elusive to interpretation by the human mind. There is a demand for 'explainable' deep learning methods to address the need for a new narrative of the machine language of the molecular sciences. This review summarizes the most prominent algorithmic concepts of explainable artificial intelligence, and dares a forecast of the future opportunities, potential applications, and remaining challenges.

arXiv:2006.16856공식 코드 · Papers with Code

A Chain Graph Interpretation of Real-World Neural Networks

The last decade has witnessed a boom of deep learning research and applications achieving state-of-the-art results in various domains. However, most advances have been established empirically, and their theoretical analysis remains lacking. One major issue is that our current interpretation of neural networks (NNs) as function approximators is too generic to support in-depth analysis. In this paper, we remedy this by proposing an alternative interpretation that identifies NNs as chain graphs (CGs) and feed-forward as an approximate inference procedure. The CG interpretation specifies the nature of each NN component within the rich theoretical framework of probabilistic graphical models, while at the same time remains general enough to cover real-world NNs with arbitrary depth, multi-branching and varied activations, as well as common structures including convolution / recurrent layers, residual block and dropout. We demonstrate with concrete examples that the CG interpretation can provide novel theoretical support and insights for various NN techniques, as well as derive new deep learning approaches such as the concept of partially collapsed feed-forward inference. It is thus a promising framework that deepens our understanding of neural networks and provides a coherent theoretical formulation for future deep learning research.

arXiv:2006.16533

Actionable Attribution Maps for Scientific Machine Learning

The scientific community has been increasingly interested in harnessing the power of deep learning to solve various domain challenges. However, despite the effectiveness in building predictive models, fundamental challenges exist in extracting actionable knowledge from the deep neural network due to their opaque nature. In this work, we propose techniques for exploring the behavior of deep learning models by injecting domain-specific actionable concepts as tunable ``knobs'' in the analysis pipeline. By incorporating the domain knowledge with generative modeling, we are not only able to better understand the behavior of these black-box models, but also provide scientists with actionable insights that can potentially lead to fundamental discoveries.

arXiv:2006.16344

Material Recognition for Automated Progress Monitoring using Deep Learning Methods

Recent advancements in Artificial intelligence, especially deep learning, has changed many fields irreversibly by introducing state of the art methods for automation. Construction monitoring has not been an exception; as a part of construction monitoring systems, material classification and recognition have drawn the attention of deep learning and machine vision researchers. However, to create production-ready systems, there is still a long path to cover. Real-world problems such as varying illuminations and reaching acceptable accuracies need to be addressed in order to create robust systems. In this paper, we have addressed these issues and reached a state of the art performance, i.e., 97.35% accuracy rate for this task. Also, a new dataset containing 1231 images of 11 classes taken from several construction sites is gathered and publicly published to help other researchers in this field.

arXiv:2006.15271공식 코드 · Papers with Code

Compressive MR Fingerprinting reconstruction with Neural Proximal Gradient iterations

Consistency of the predictions with respect to the physical forward model is pivotal for reliably solving inverse problems. This consistency is mostly un-controlled in the current end-to-end deep learning methodologies proposed for the Magnetic Resonance Fingerprinting (MRF) problem. To address this, we propose ProxNet, a learned proximal gradient descent framework that directly incorporates the forward acquisition and Bloch dynamic models within a recurrent learning mechanism. The ProxNet adopts a compact neural proximal model for de-aliasing and quantitative inference, that can be flexibly trained on scarce MRF training datasets. Our numerical experiments show that the ProxNet can achieve a superior quantitative inference accuracy, much smaller storage requirement, and a comparable runtime to the recent deep learning MRF baselines, while being much faster than the dictionary matching schemes. Code has been released at https://github.com/edongdongchen/PGD-Net.

arXiv:2006.15771

Active Ensemble Deep Learning for Polarimetric Synthetic Aperture Radar Image Classification

Although deep learning has achieved great success in image classification tasks, its performance is subject to the quantity and quality of training samples. For classification of polarimetric synthetic aperture radar (PolSAR) images, it is nearly impossible to annotate the images from visual interpretation. Therefore, it is urgent for remote sensing scientists to develop new techniques for PolSAR image classification under the condition of very few training samples. In this letter, we take the advantage of active learning and propose active ensemble deep learning (AEDL) for PolSAR image classification. We first show that only 35\% of the predicted labels of a deep learning model's snapshots near its convergence were exactly the same. The disagreement between snapshots is non-negligible. From the perspective of multiview learning, the snapshots together serve as a good committee to evaluate the importance of unlabeled instances. Using the snapshots committee to give out the informativeness of unlabeled data, the proposed AEDL achieved better performance on two real PolSAR images compared with standard active learning strategies. It achieved the same classification accuracy with only 86% and 55% of the training samples compared with breaking ties active learning and random selection for the Flevoland dataset.

arXiv:2006.15766공식 코드 · Papers with Code

Heteroskedastic and Imbalanced Deep Learning with Adaptive Regularization

Real-world large-scale datasets are heteroskedastic and imbalanced -- labels have varying levels of uncertainty and label distributions are long-tailed. Heteroskedasticity and imbalance challenge deep learning algorithms due to the difficulty of distinguishing among mislabeled, ambiguous, and rare examples. Addressing heteroskedasticity and imbalance simultaneously is under-explored. We propose a data-dependent regularization technique for heteroskedastic datasets that regularizes different regions of the input space differently. Inspired by the theoretical derivation of the optimal regularization strength in a one-dimensional nonparametric classification setting, our approach adaptively regularizes the data points in higher-uncertainty, lower-density regions more heavily. We test our method on several benchmark tasks, including a real-world heteroskedastic and imbalanced dataset, WebVision. Our experiments corroborate our theory and demonstrate a significant improvement over other methods in noise-robust deep learning.

arXiv:2006.16161

A Two-step Surface-based 3D Deep Learning Pipeline for Segmentation of Intracranial Aneurysms

The exact shape of intracranial aneurysms is critical in medical diagnosis and surgical planning. While voxel-based deep learning frameworks have been proposed for this segmentation task, their performance remains limited. In this study, we offer a two-step surface-based deep learning pipeline that achieves significantly higher performance. Our proposed model takes a surface model of entire principal brain arteries containing aneurysms as input and returns aneurysms surfaces as output. A user first generates a surface model by manually specifying multiple thresholds for time-of-flight magnetic resonance angiography images. The system then samples small surface fragments from the entire brain arteries and classifies the surface fragments according to whether aneurysms are present using a point-based deep learning network (PointNet++). Finally, the system applies surface segmentation (SO-Net) to surface fragments containing aneurysms. We conduct a direct comparison of segmentation performance by counting voxels between the proposed surface-based framework and the existing voxel-based method, in which our framework achieves a much higher dice similarity coefficient score (72%) than the prior approach (46%).

arXiv:2006.16148공식 코드 · Papers with Code

Large Deformation Diffeomorphic Image Registration with Laplacian Pyramid Networks

Deep learning-based methods have recently demonstrated promising results in deformable image registration for a wide range of medical image analysis tasks. However, existing deep learning-based methods are usually limited to small deformation settings, and desirable properties of the transformation including bijective mapping and topology preservation are often being ignored by these approaches. In this paper, we propose a deep Laplacian Pyramid Image Registration Network, which can solve the image registration optimization problem in a coarse-to-fine fashion within the space of diffeomorphic maps. Extensive quantitative and qualitative evaluations on two MR brain scan datasets show that our method outperforms the existing methods by a significant margin while maintaining desirable diffeomorphic properties and promising registration speed.

arXiv:2006.14651

Influence Functions in Deep Learning Are Fragile

Influence functions approximate the effect of training samples in test-time predictions and have a wide variety of applications in machine learning interpretability and uncertainty estimation. A commonly-used (first-order) influence function can be implemented efficiently as a post-hoc method requiring access only to the gradients and Hessian of the model. For linear models, influence functions are well-defined due to the convexity of the underlying loss function and are generally accurate even across difficult settings where model changes are fairly large such as estimating group influences. Influence functions, however, are not well-understood in the context of deep learning with non-convex loss functions. In this paper, we provide a comprehensive and large-scale empirical study of successes and failures of influence functions in neural network models trained on datasets such as Iris, MNIST, CIFAR-10 and ImageNet. Through our extensive experiments, we show that the network architecture, its depth and width, as well as the extent of model parameterization and regularization techniques have strong effects in the accuracy of influence functions. In particular, we find that (i) influence estimates are fairly accurate for shallow networks, while for deeper networks the estimates are often erroneous; (ii) for certain network architectures and datasets, training with weight-decay regularization is important to get high-quality influence estimates; and (iii) the accuracy of influence estimates can vary significantly depending on the examined test points. These results suggest that in general influence functions in deep learning are fragile and call for developing improved influence estimation methods to mitigate these issues in non-convex setups.

arXiv:2006.11395

Deep Learning Based Single Sample Per Person Face Recognition: A Survey

Face recognition has long been an active research area in the field of artificial intelligence, particularly since the rise of deep learning in recent years. In some practical situations, each identity has only a single sample available for training. Face recognition under this situation is referred to as single sample face recognition and poses significant challenges to the effective training of deep models. Therefore, in recent years, researchers have attempted to unleash more potential of deep learning and improve the model recognition performance in the single sample situation. While several comprehensive surveys have been conducted on traditional single sample face recognition approaches, emerging deep learning based methods are rarely involved in these reviews. Accordingly, we focus on the deep learning-based methods in this paper, classifying them into virtual sample methods and generic learning methods. In the former category, virtual images or virtual features are generated to benefit the training of the deep model. In the latter one, additional multi-sample generic sets are used. There are three types of generic learning methods: combining traditional methods and deep features, improving the loss function, and improving network structure, all of which are covered in our analysis. Moreover, we review face datasets that have been commonly used for evaluating single sample face recognition models and go on to compare the results of different types of models. Additionally, we discuss problems with existing single sample face recognition methods, including identity information preservation in virtual sample methods, domain adaption in generic learning methods. Furthermore, we regard developing unsupervised methods is a promising future direction, and point out that the semantic gap as an important issue that needs to be further considered.

arXiv:2006.14319공식 코드 · Papers with Code

Deep Learning for Cornea Microscopy Blind Deblurring

The goal of this project is to build a deep-learning solution that deblurs cornea scans, used for medical examination. The spherical shape of the eye prevents ophtamologist from having completely sharp image. Provided with a stack of corneas from confocal images, our approach is to build a model that performs an upscaling of the images using an SR (Super Resolution) Network.

arXiv:2006.14170

Towards Differentially Private Text Representations

Most deep learning frameworks require users to pool their local data or model updates to a trusted server to train or maintain a global model. The assumption of a trusted server who has access to user information is ill-suited in many applications. To tackle this problem, we develop a new deep learning framework under an untrusted server setting, which includes three modules: (1) embedding module, (2) randomization module, and (3) classifier module. For the randomization module, we propose a novel local differentially private (LDP) protocol to reduce the impact of privacy parameter $\epsilon$ on accuracy, and provide enhanced flexibility in choosing randomization probabilities for LDP. Analysis and experiments show that our framework delivers comparable or even better performance than the non-private framework and existing LDP protocols, demonstrating the advantages of our LDP protocol.

arXiv:2006.14580

Backdoor Attacks Against Deep Learning Systems in the Physical World

Backdoor attacks embed hidden malicious behaviors into deep learning models, which only activate and cause misclassifications on model inputs containing a specific trigger. Existing works on backdoor attacks and defenses, however, mostly focus on digital attacks that use digitally generated patterns as triggers. A critical question remains unanswered: can backdoor attacks succeed using physical objects as triggers, thus making them a credible threat against deep learning systems in the real world? We conduct a detailed empirical study to explore this question for facial recognition, a critical deep learning task. Using seven physical objects as triggers, we collect a custom dataset of 3205 images of ten volunteers and use it to study the feasibility of physical backdoor attacks under a variety of real-world conditions. Our study reveals two key findings. First, physical backdoor attacks can be highly successful if they are carefully configured to overcome the constraints imposed by physical objects. In particular, the placement of successful triggers is largely constrained by the target model's dependence on key facial features. Second, four of today's state-of-the-art defenses against (digital) backdoors are ineffective against physical backdoors, because the use of physical objects breaks core assumptions used to construct these defenses. Our study confirms that (physical) backdoor attacks are not a hypothetical phenomenon but rather pose a serious real-world threat to critical classification tasks. We need new and more robust defenses against backdoors in the physical world.

arXiv:2006.13878

Effective Elastic Scaling of Deep Learning Workloads

The increased use of deep learning (DL) in academia, government and industry has, in turn, led to the popularity of on-premise and cloud-hosted deep learning platforms, whose goals are to enable organizations utilize expensive resources effectively, and to share said resources among multiple teams in a fair and effective manner. In this paper, we examine the elastic scaling of Deep Learning (DL) jobs over large-scale training platforms and propose a novel resource allocation strategy for DL training jobs, resulting in improved job run time performance as well as increased cluster utilization. We begin by analyzing DL workloads and exploit the fact that DL jobs can be run with a range of batch sizes without affecting their final accuracy. We formulate an optimization problem that explores a dynamic batch size allocation to individual DL jobs based on their scaling efficiency, when running on multiple nodes. We design a fast dynamic programming based optimizer to solve this problem in real-time to determine jobs that can be scaled up/down, and use this optimizer in an autoscaler to dynamically change the allocated resources and batch sizes of individual DL jobs. We demonstrate empirically that our elastic scaling algorithm can complete up to $\approx 2 \times$ as many jobs as compared to a strong baseline algorithm that also scales the number of GPUs but does not change the batch size. We also demonstrate that the average completion time with our algorithm is up to $\approx 10 \times$ faster than that of the baseline.

arXiv:2006.13215

Traffic congestion anomaly detection and prediction using deep learning

Congestion prediction represents a major priority for traffic management centres around the world to ensure timely incident response handling. The increasing amounts of generated traffic data have been used to train machine learning predictors for traffic, however, this is a challenging task due to inter-dependencies of traffic flow both in time and space. Recently, deep learning techniques have shown significant prediction improvements over traditional models, however, open questions remain around their applicability, accuracy and parameter tuning. This paper brings two contributions in terms of: 1) applying an outlier detection an anomaly adjustment method based on incoming and historical data streams, and 2) proposing an advanced deep learning framework for simultaneously predicting the traffic flow, speed and occupancy on a large number of monitoring stations along a highly circulated motorway in Sydney, Australia, including exit and entry loop count stations, and over varying training and prediction time horizons. The spatial and temporal features extracted from the 36.34 million data points are used in various deep learning architectures that exploit their spatial structure (convolutional neuronal networks), their temporal dynamics (recurrent neuronal networks), or both through a hybrid spatio-temporal modelling (CNN-LSTM). We show that our deep learning models consistently outperform traditional methods, and we conduct a comparative analysis of the optimal time horizon of historical data required to predict traffic flow at different time points in the future. Lastly, we prove that the anomaly adjustment method brings significant improvements to using deep learning in both time and space.

arXiv:2006.13554공식 코드 · Papers with Code

Normalized Loss Functions for Deep Learning with Noisy Labels

Robust loss functions are essential for training accurate deep neural networks (DNNs) in the presence of noisy (incorrect) labels. It has been shown that the commonly used Cross Entropy (CE) loss is not robust to noisy labels. Whilst new loss functions have been designed, they are only partially robust. In this paper, we theoretically show by applying a simple normalization that: any loss can be made robust to noisy labels. However, in practice, simply being robust is not sufficient for a loss function to train accurate DNNs. By investigating several robust loss functions, we find that they suffer from a problem of underfitting. To address this, we propose a framework to build robust loss functions called Active Passive Loss (APL). APL combines two robust loss functions that mutually boost each other. Experiments on benchmark datasets demonstrate that the family of new loss functions created by our APL framework can consistently outperform state-of-the-art methods by large margins, especially under large noise rates such as 60% or 80% incorrect labels.

arXiv:2006.12915

Deep Attentive Wasserstein Generative Adversarial Networks for MRI Reconstruction with Recurrent Context-Awareness

The performance of traditional compressive sensing-based MRI (CS-MRI) reconstruction is affected by its slow iterative procedure and noise-induced artefacts. Although many deep learning-based CS-MRI methods have been proposed to mitigate the problems of traditional methods, they have not been able to achieve more robust results at higher acceleration factors. Most of the deep learning-based CS-MRI methods still can not fully mine the information from the k-space, which leads to unsatisfactory results in the MRI reconstruction. In this study, we propose a new deep learning-based CS-MRI reconstruction method to fully utilise the relationship among sequential MRI slices by coupling Wasserstein Generative Adversarial Networks (WGAN) with Recurrent Neural Networks. Further development of an attentive unit enables our model to reconstruct more accurate anatomical structures for the MRI data. By experimenting on different MRI datasets, we have demonstrated that our method can not only achieve better results compared to the state-of-the-arts but can also effectively reduce residual noise generated during the reconstruction process.