PyTorchKR Papers

최신 논문

총 8,922편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2110.00459

Characterizing Concurrency Mechanisms for NVIDIA GPUs under Deep Learning Workloads

We investigate the performance of the concurrency mechanisms available on NVIDIA's new Ampere GPU microarchitecture under deep learning training and inference workloads. In contrast to previous studies that treat the GPU as a black box, we examine scheduling at the microarchitectural level. We find that the lack of fine-grained preemption mechanisms, robust task prioritization options, and contention-aware thread block placement policies limits the effectiveness of NVIDIA's concurrency mechanisms. In summary, the sequential nature of deep learning workloads and their fluctuating resource requirements and kernel runtimes make executing such workloads while maintaining consistently high utilization and low, predictable turnaround times difficult on current NVIDIA hardware.

arXiv:2110.00199공식 코드 · Papers with Code

Perturbated Gradients Updating within Unit Space for Deep Learning

In deep learning, optimization plays a vital role. By focusing on image classification, this work investigates the pros and cons of the widely used optimizers, and proposes a new optimizer: Perturbated Unit Gradient Descent (PUGD) algorithm with extending normalized gradient operation in tensor within perturbation to update in unit space. Via a set of experiments and analyses, we show that PUGD is locally bounded updating, which means the updating from time to time is controlled. On the other hand, PUGD can push models to a flat minimum, where the error remains approximately constant, not only because of the nature of avoiding stationary points in gradient normalization but also by scanning sharpness in the unit ball. From a series of rigorous experiments, PUGD helps models to gain a state-of-the-art Top-1 accuracy in Tiny ImageNet and competitive performances in CIFAR- {10, 100}. We open-source our code at link: https://github.com/hanktseng131415go/PUGD.

arXiv:2110.00111

Deep Learning-based Action Detection in Untrimmed Videos: A Survey

Understanding human behavior and activity facilitates advancement of numerous real-world applications, and is critical for video analysis. Despite the progress of action recognition algorithms in trimmed videos, the majority of real-world videos are lengthy and untrimmed with sparse segments of interest. The task of temporal activity detection in untrimmed videos aims to localize the temporal boundary of actions and classify the action categories. Temporal activity detection task has been investigated in full and limited supervision settings depending on the availability of action annotations. This paper provides an extensive overview of deep learning-based algorithms to tackle temporal action detection in untrimmed videos with different supervision levels including fully-supervised, weakly-supervised, unsupervised, self-supervised, and semi-supervised. In addition, this paper also reviews advances in spatio-temporal action detection where actions are localized in both temporal and spatial dimensions. Moreover, the commonly used action detection benchmark datasets and evaluation metrics are described, and the performance of the state-of-the-art methods are compared. Finally, real-world applications of temporal action detection in untrimmed videos and a set of future directions are discussed.

arXiv:2109.15144

A Review of Text Style Transfer using Deep Learning

Style is an integral component of a sentence indicated by the choice of words a person makes. Different people have different ways of expressing themselves, however, they adjust their speaking and writing style to a social context, an audience, an interlocutor or the formality of an occasion. Text style transfer is defined as a task of adapting and/or changing the stylistic manner in which a sentence is written, while preserving the meaning of the original sentence. A systematic review of text style transfer methodologies using deep learning is presented in this paper. We point out the technological advances in deep neural networks that have been the driving force behind current successes in the fields of natural language understanding and generation. The review is structured around two key stages in the text style transfer process, namely, representation learning and sentence generation in a new style. The discussion highlights the commonalities and differences between proposed solutions as well as challenges and opportunities that are expected to direct and foster further research in the field.

arXiv:2109.14737

Unlocking the potential of deep learning for marine ecology: overview, applications, and outlook

The deep learning revolution is touching all scientific disciplines and corners of our lives as a means of harnessing the power of big data. Marine ecology is no exception. These new methods provide analysis of data from sensors, cameras, and acoustic recorders, even in real time, in ways that are reproducible and rapid. Off-the-shelf algorithms can find, count, and classify species from digital images or video and detect cryptic patterns in noisy data. Using these opportunities requires collaboration across ecological and data science disciplines, which can be challenging to initiate. To facilitate these collaborations and promote the use of deep learning towards ecosystem-based management of the sea, this paper aims to bridge the gap between marine ecologists and computer scientists. We provide insight into popular deep learning approaches for ecological data analysis in plain language, focusing on the techniques of supervised learning with deep neural networks, and illustrate challenges and opportunities through established and emerging applications of deep learning to marine ecology. We use established and future-looking case studies on plankton, fishes, marine mammals, pollution, and nutrient cycling that involve object detection, classification, tracking, and segmentation of visualized data. We conclude with a broad outlook of the field's opportunities and challenges, including potential technological advances and issues with managing complex data sets.

arXiv:2109.14545공식 코드 · Papers with Code

Activation Functions in Deep Learning: A Comprehensive Survey and Benchmark

Neural networks have shown tremendous growth in recent years to solve numerous problems. Various types of neural networks have been introduced to deal with different types of problems. However, the main goal of any neural network is to transform the non-linearly separable input data into more linearly separable abstract features using a hierarchy of layers. These layers are combinations of linear and nonlinear functions. The most popular and common non-linearity layers are activation functions (AFs), such as Logistic Sigmoid, Tanh, ReLU, ELU, Swish and Mish. In this paper, a comprehensive overview and survey is presented for AFs in neural networks for deep learning. Different classes of AFs such as Logistic Sigmoid and Tanh based, ReLU based, ELU based, and Learning based are covered. Several characteristics of AFs such as output range, monotonicity, and smoothness are also pointed out. A performance comparison is also performed among 18 state-of-the-art AFs with different networks on different types of data. The insights of AFs are presented to benefit the researchers for doing further research and practitioners to select among different choices. The code used for experimental comparison is released at: \url{https://github.com/shivram1987/ActivationFunctions}.

arXiv:2109.14530공식 코드 · Papers with Code

Deep Spatio-Temporal Wind Power Forecasting

Wind power forecasting has drawn increasing attention among researchers as the consumption of renewable energy grows. In this paper, we develop a deep learning approach based on encoder-decoder structure. Our model forecasts wind power generated by a wind turbine using its spatial location relative to other turbines and historical wind speed data. In this way, we effectively integrate spatial dependency and temporal trends to make turbine-specific predictions. The advantages of our method over existing work can be summarized as 1) it directly predicts wind power based on historical wind speed, without the need for prediction of wind speed first, and then using a transformation; 2) it can effectively capture long-term dependency 3) our model is more scalable and efficient compared with other deep learning based methods. We demonstrate the efficacy of our model on the benchmark real-world datasets.

arXiv:2109.14433

Multi-loss ensemble deep learning for chest X-ray classification

Medical images commonly exhibit multiple abnormalities. Predicting them requires multi-class classifiers whose training and desired reliable performance can be affected by a combination of factors, such as, dataset size, data source, distribution, and the loss function used to train the deep neural networks. Currently, the cross-entropy loss remains the de-facto loss function for training deep learning classifiers. This loss function, however, asserts equal learning from all classes, leading to a bias toward the majority class. In this work, we benchmark various state-of-the-art loss functions that are suitable for multi-class classification, critically analyze model performance, and propose improved loss functions. We select a pediatric chest X-ray (CXR) dataset that includes images with no abnormality (normal), and those exhibiting manifestations consistent with bacterial and viral pneumonia. We construct prediction-level and model-level ensembles, respectively, to improve classification performance. Our results show that compared to the individual models and the state-of-the-art literature, the weighted averaging of the predictions for top-3 and top-5 model-level ensembles delivered significantly superior classification performance (p < 0.05) in terms of MCC (0.9068, 95% confidence interval (0.8839, 0.9297)) metric. Finally, we performed localization studies to interpret model behaviors to visualize and confirm that the individual models and ensembles learned meaningful features and highlighted disease manifestations.

arXiv:2109.14335공식 코드 · Papers with Code

A Systematic Survey of Deep Learning-based Single-Image Super-Resolution

Single-image super-resolution (SISR) is an important task in image processing, which aims to enhance the resolution of imaging systems. Recently, SISR has made a huge leap and has achieved promising results with the help of deep learning (DL). In this survey, we give an overview of DL-based SISR methods and group them according to their design targets. Specifically, we first introduce the problem definition, research background, and the significance of SISR. Secondly, we introduce some related works, including benchmark datasets, upsampling methods, optimization objectives, and image quality assessment methods. Thirdly, we provide a detailed investigation of SISR and give some domain-specific applications of it. Fourthly, we present the reconstruction results of some classic SISR methods to intuitively know their performance. Finally, we discuss some issues that still exist in SISR and summarize some new trends and future directions. This is an exhaustive survey of SISR, which can help researchers better understand SISR and inspire more exciting research in this field. An investigation project for SISR is provided at https://github.com/CV-JunchengLi/SISR-Survey.

arXiv:2109.13237

DOODLER: Determining Out-Of-Distribution Likelihood from Encoder Reconstructions

Deep Learning models possess two key traits that, in combination, make their use in the real world a risky prospect. One, they do not typically generalize well outside of the distribution for which they were trained, and two, they tend to exhibit confident behavior regardless of whether or not they are producing meaningful outputs. While Deep Learning possesses immense power to solve realistic, high-dimensional problems, these traits in concert make it difficult to have confidence in their real-world applications. To overcome this difficulty, the task of Out-Of-Distribution (OOD) Detection has been defined, to determine when a model has received an input from outside of the distribution for which it is trained to operate. This paper introduces and examines a novel methodology, DOODLER, for OOD Detection, which directly leverages the traits which result in its necessity. By training a Variational Auto-Encoder (VAE) on the same data as another Deep Learning model, the VAE learns to accurately reconstruct In-Distribution (ID) inputs, but not to reconstruct OOD inputs, meaning that its failure state can be used to perform OOD Detection. Unlike other work in the area, DOODLER requires only very weak assumptions about the existence of an OOD dataset, allowing for more realistic application. DOODLER also enables pixel-wise segmentations of input images by OOD likelihood, and experimental results show that it matches or outperforms methodologies that operate under the same constraints.

arXiv:2109.13531

Multi-Semantic Image Recognition Model and Evaluating Index for explaining the deep learning models

Although deep learning models are powerful among various applications, most deep learning models are still a black box, lacking verifiability and interpretability, which means the decision-making process that human beings cannot understand. Therefore, how to evaluate deep neural networks with explanations is still an urgent task. In this paper, we first propose a multi-semantic image recognition model, which enables human beings to understand the decision-making process of the neural network. Then, we presents a new evaluation index, which can quantitatively assess the model interpretability. We also comprehensively summarize the semantic information that affects the image classification results in the judgment process of neural networks. Finally, this paper also exhibits the relevant baseline performance with current state-of-the-art deep learning models.

arXiv:2109.13119

Plane-Wave Ultrasound Beamforming: A Deep Learning Approach

Medical ultrasound provides images which are the spatial map of the tissue echogenicity. Unfortunately, an ultrasound image is a low-quality version of the expected Tissue Reflectivity Function (TRF) mainly due to the non-ideal Point Spread Function (PSF) of the imaging system. This paper presents a novel beamforming approach based on deep learning to get closer to the ideal PSF in Plane-Wave Imaging (PWI). The proposed approach is designed to reconstruct the desired TRF from echo traces acquired by transducer elements using only a single plane-wave transmission. In this approach, first, an ideal model for the TRF is introduced by setting the imaging PSF as a sharp Gaussian function. Then, a mapping function between the pre-beamformed Radio-Frequency (RF) channel data and the proposed TRF is constructed using deep learning. Network architecture contains multi-resolution decomposition and reconstruction using wavelet transform for effective recovery of high-frequency content of the desired TRF. Inspired by curriculum learning, we exploit step by step training from coarse (mean square error) to fine ($\ell_{0.2}$) loss functions. The proposed method is trained on a large number of simulation ultrasound data with the ground-truth echogenicity map extracted from real photographic images. The performance of the trained network is evaluated on the publicly available simulation and \textit{in vivo} test data without any further fine-tuning. Simulation test results confirm that the proposed method reconstructs images with a high quality in terms of resolution and contrast, which are also visually similar to the proposed ground-truth image. Furthermore, \textit{in vivo} results show that the trained mapping function preserves its performance in the new domain. Therefore, the proposed approach maintains high resolution, contrast, and framerate simultaneously.

arXiv:2109.12404

Deep learning tackles single-cell analysis A survey of deep learning for scRNA-seq analysis

Since its selection as the method of the year in 2013, single-cell technologies have become mature enough to provide answers to complex research questions. With the growth of single-cell profiling technologies, there has also been a significant increase in data collected from single-cell profilings, resulting in computational challenges to process these massive and complicated datasets. To address these challenges, deep learning (DL) is positioning as a competitive alternative for single-cell analyses besides the traditional machine learning approaches. Here we present a processing pipeline of single-cell RNA-seq data, survey a total of 25 DL algorithms and their applicability for a specific step in the processing pipeline. Specifically, we establish a unified mathematical representation of all variational autoencoder, autoencoder, and generative adversarial network models, compare the training strategies and loss functions for these models, and relate the loss functions of these models to specific objectives of the data processing step. Such presentation will allow readers to choose suitable algorithms for their particular objective at each step in the pipeline. We envision that this survey will serve as an important information portal for learning the application of DL for scRNA-seq analysis and inspire innovative use of DL to address a broader range of new challenges in emerging multi-omics and spatial single-cell sequencing.

arXiv:2109.12520

Entity Linking Meets Deep Learning: Techniques and Solutions

Entity linking (EL) is the process of linking entity mentions appearing in web text with their corresponding entities in a knowledge base. EL plays an important role in the fields of knowledge engineering and data mining, underlying a variety of downstream applications such as knowledge base population, content analysis, relation extraction, and question answering. In recent years, deep learning (DL), which has achieved tremendous success in various domains, has also been leveraged in EL methods to surpass traditional machine learning based methods and yield the state-of-the-art performance. In this survey, we present a comprehensive review and analysis of existing DL based EL methods. First of all, we propose a new taxonomy, which organizes existing DL based EL methods using three axes: embedding, feature, and algorithm. Then we systematically survey the representative EL methods along the three axes of the taxonomy. Later, we introduce ten commonly used EL data sets and give a quantitative performance analysis of DL based EL methods over these data sets. Finally, we discuss the remaining limitations of existing methods and highlight some promising future directions.

arXiv:2109.13012

Federated Deep Learning with Bayesian Privacy

Federated learning (FL) aims to protect data privacy by cooperatively learning a model without sharing private data among users. For Federated Learning of Deep Neural Network with billions of model parameters, existing privacy-preserving solutions are unsatisfactory. Homomorphic encryption (HE) based methods provide secure privacy protections but suffer from extremely high computational and communication overheads rendering it almost useless in practice . Deep learning with Differential Privacy (DP) was implemented as a practical learning algorithm at a manageable cost in complexity. However, DP is vulnerable to aggressive Bayesian restoration attacks as disclosed in the literature and demonstrated in experimental results of this work. To address the aforementioned perplexity, we propose a novel Bayesian Privacy (BP) framework which enables Bayesian restoration attacks to be formulated as the probability of reconstructing private data from observed public information. Specifically, the proposed BP framework accurately quantifies privacy loss by Kullback-Leibler (KL) Divergence between the prior distribution about the privacy data and the posterior distribution of restoration private data conditioning on exposed information}. To our best knowledge, this Bayesian Privacy analysis is the first to provides theoretical justification of secure privacy-preserving capabilities against Bayesian restoration attacks. As a concrete use case, we demonstrate that a novel federated deep learning method using private passport layers is able to simultaneously achieve high model performance, privacy-preserving capability and low computational complexity. Theoretical analysis is in accordance with empirical measurements of information leakage extensively experimented with a variety of DNN networks on image classification MNIST, CIFAR10, and CIFAR100 datasets.

arXiv:2109.12970

Deep Learning Based Resource Assignment for Wireless Networks

This paper studies a deep learning approach for binary assignment problems in wireless networks, which identifies binary variables for permutation matrices. This poses challenges in designing a structure of a neural network and its training strategies for generating feasible assignment solutions. To this end, this paper develop a new Sinkhorn neural network which learns a non-convex projection task onto a set of permutation matrices. An unsupervised training algorithm is proposed where the Sinkhorn neural network can be applied to network assignment problems. Numerical results demonstrate the effectiveness of the proposed method in various network scenarios.

arXiv:2109.12894공식 코드 · Papers with Code

Training Spiking Neural Networks Using Lessons From Deep Learning

The brain is the perfect place to look for inspiration to develop more efficient neural networks. The inner workings of our synapses and neurons provide a glimpse at what the future of deep learning might look like. This paper serves as a tutorial and perspective showing how to apply the lessons learnt from several decades of research in deep learning, gradient descent, backpropagation and neuroscience to biologically plausible spiking neural neural networks. We also explore the delicate interplay between encoding data as spikes and the learning process; the challenges and solutions of applying gradient-based learning to spiking neural networks (SNNs); the subtle link between temporal backpropagation and spike timing dependent plasticity, and how deep learning might move towards biologically plausible online learning. Some ideas are well accepted and commonly used amongst the neuromorphic engineering community, while others are presented or justified for the first time here. The fields of deep learning and spiking neural networks evolve very rapidly. We endeavour to treat this document as a 'dynamic' manuscript that will continue to be updated as the common practices in training SNNs also change. A series of companion interactive tutorials complementary to this paper using our Python package, snnTorch, are also made available. See https://snntorch.readthedocs.io/en/latest/tutorials/index.html .

arXiv:2109.12398

Channel State Information Based Localization with Deep Learning

Localization is one of the most important problems in various fields such as robotics and wireless communications. For instance, Unmanned Aerial Vehicles (UAVs) require the information of the position precisely for an adequate control strategy. This problem is handled very efficiently with integrated GPS units for outdoor applications. However, indoor applications require special treatment due to the unavailability of GPS signals. Another aspect of mobile robots such as UAVs is that there is constant wireless communication between the mobile robot and a computational unit. This communication is mainly done for obtaining telemetry information or computation of control actions directly. The responsible integrated units for this transmission are commercial wireless communication chipsets. These units on the receiver side are responsible for getting rid of the diverse effects of the communication channel with various mathematical techniques. These techniques mainly require the Channel State Information (CSI) of the current channel to compensate the channel itself. After the compensation, the chipset has nothing to do with CSI. However, the locations of both the transmitter and receiver have a direct impact on CSI. Even though CSI contains such rich information about the environment, the accessibility of these data is blocked by the commercial wireless chipsets since they are manufactured to provide only the processed information data bits to the user. However, with the IEEE 802.11n standardization, certain chipsets provide access to CSI. Therefore, CSI data became processible and integrable to localization schemes. In this project, a test environment was constructed for the localization task. Two routers with proper chipsets were assigned as transmitter and receiver. They were operationalized for the CSI data collection. Lastly, these data were processed with various deep learning models.

arXiv:2109.13157

Visual Anomaly Detection for Images: A Survey

Visual anomaly detection is an important and challenging problem in the field of machine learning and computer vision. This problem has attracted a considerable amount of attention in relevant research communities. Especially in recent years, the development of deep learning has sparked an increasing interest in the visual anomaly detection problem and brought a great variety of novel methods. In this paper, we provide a comprehensive survey of the classical and deep learning-based approaches for visual anomaly detection in the literature. We group the relevant approaches in view of their underlying principles and discuss their assumptions, advantages, and disadvantages carefully. We aim to help the researchers to understand the common principles of visual anomaly detection approaches and identify promising research directions in this field.

arXiv:2109.12845

Bayesian deep learning of affordances from RGB images

Autonomous agents, such as robots or intelligent devices, need to understand how to interact with objects and its environment. Affordances are defined as the relationships between an agent, the objects, and the possible future actions in the environment. In this paper, we present a Bayesian deep learning method to predict the affordances available in the environment directly from RGB images. Based on previous work on socially accepted affordances, our model is based on a multiscale CNN that combines local and global information from the object and the full image. However, previous works assume a deterministic model, but uncertainty quantification is fundamental for robust detection, affordance-based reason, continual learning, etc. Our Bayesian model is able to capture both the aleatoric uncertainty from the scene and the epistemic uncertainty associated with the model and previous learning process. For comparison, we estimate the uncertainty using two state-of-the-art techniques: Monte Carlo dropout and deep ensembles. We also compare different types of CNN encoders for feature extraction. We have performed several experiments on an affordance database on socially acceptable behaviours and we have shown improved performance compared with previous works. Furthermore, the uncertainty estimation is consistent with the the type of objects and scenarios. Our results show a marginal better performance of deep ensembles, compared to MC-dropout on the Brier score and the Expected Calibration Error.