PyTorchKR Papers

최신 논문

총 2,063편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:1811.12488공식 코드 · Papers with Code

Leveraging Deep Stein's Unbiased Risk Estimator for Unsupervised X-ray Denoising

Among the plethora of techniques devised to curb the prevalence of noise in medical images, deep learning based approaches have shown the most promise. However, one critical limitation of these deep learning based denoisers is the requirement of high-quality noiseless ground truth images that are difficult to obtain in many medical imaging applications such as X-rays. To circumvent this issue, we leverage recently proposed approach of [7] that incorporates Stein's Unbiased Risk Estimator (SURE) to train a deep convolutional neural network without requiring denoised ground truth X-ray data. Our experimental results demonstrate the effectiveness of SURE based approach for denoising X-ray images.

arXiv:1811.12929공식 코드 · Papers with Code

Online Abstraction with MDP Homomorphisms for Deep Learning

Abstraction of Markov Decision Processes is a useful tool for solving complex problems, as it can ignore unimportant aspects of an environment, simplifying the process of learning an optimal policy. In this paper, we propose a new algorithm for finding abstract MDPs in environments with continuous state spaces. It is based on MDP homomorphisms, a structure-preserving mapping between MDPs. We demonstrate our algorithm's ability to learn abstractions from collected experience and show how to reuse the abstractions to guide exploration in new tasks the agent encounters. Our novel task transfer method outperforms baselines based on a deep Q-network in the majority of our experiments. The source code is at https://github.com/ondrejba/aamas_19.

arXiv:1811.11880공식 코드 · Papers with Code

Predicting the Computational Cost of Deep Learning Models

Deep learning is rapidly becoming a go-to tool for many artificial intelligence problems due to its ability to outperform other approaches and even humans at many problems. Despite its popularity we are still unable to accurately predict the time it will take to train a deep learning network to solve a given problem. This training time can be seen as the product of the training time per epoch and the number of epochs which need to be performed to reach the desired level of accuracy. Some work has been carried out to predict the training time for an epoch -- most have been based around the assumption that the training time is linearly related to the number of floating point operations required. However, this relationship is not true and becomes exacerbated in cases where other activities start to dominate the execution time. Such as the time to load data from memory or loss of performance due to non-optimal parallel execution. In this work we propose an alternative approach in which we train a deep learning network to predict the execution time for parts of a deep learning network. Timings for these individual parts can then be combined to provide a prediction for the whole execution time. This has advantages over linear approaches as it can model more complex scenarios. But, also, it has the ability to predict execution times for scenarios unseen in the training data. Therefore, our approach can be used not only to infer the execution time for a batch, or entire epoch, but it can also support making a well-informed choice for the appropriate hardware and model.

arXiv:1811.11405공식 코드 · Papers with Code

Spectral Feature Transformation for Person Re-identification

With the surge of deep learning techniques, the field of person re-identification has witnessed rapid progress in recent years. Deep learning based methods focus on learning a feature space where samples are clustered compactly according to their corresponding identities. Most existing methods rely on powerful CNNs to transform the samples individually. In contrast, we propose to consider the sample relations in the transformation. To achieve this goal, we incorporate spectral clustering technique into CNN. We derive a novel module named Spectral Feature Transformation and seamlessly integrate it into existing CNN pipeline with negligible cost,which makes our method enjoy the best of two worlds. Empirical studies show that the proposed approach outperforms previous state-of-the-art methods on four public benchmarks by a considerable margin without bells and whistles.

arXiv:1811.09651공식 코드 · Papers with Code

A New Cervical Cytology Dataset for Nucleus Detection and Image Classification (Cervix93) and Methods for Cervical Nucleus Detection

Analyzing Pap cytology slides is an important tasks in detecting and grading precancerous and cancerous cervical cancer stages. Processing cytology images usually involve segmenting nuclei and overlapping cells. We introduce a cervical cytology dataset that can be used to evaluate nucleus detection, as well as image classification methods in the cytology image processing area. This dataset contains 93 real image stacks with their grade labels and manually annotated nuclei within images. We also present two methods: a baseline method based on a previously proposed approach, and a deep learning method, and compare their results with other state-of-the-art methods. Both the baseline method and the deep learning method outperform other state-of-the-art methods by significant margins. Along with the dataset, we publicly make the evaluation code and the baseline method available to download for further benchmarking.

arXiv:1811.10167공식 코드 · Papers with Code

LSICC: A Large Scale Informal Chinese Corpus

Deep learning based natural language processing model is proven powerful, but need large-scale dataset. Due to the significant gap between the real-world tasks and existing Chinese corpus, in this paper, we introduce a large-scale corpus of informal Chinese. This corpus contains around 37 million book reviews and 50 thousand netizen's comments to the news. We explore the informal words frequencies of the corpus and show the difference between our corpus and the existing ones. The corpus can be further used to train deep learning based natural language processing tasks such as Chinese word segmentation, sentiment analysis.

arXiv:1811.10052공식 코드 · Papers with Code

An overview of deep learning in medical imaging focusing on MRI

What has happened in machine learning lately, and what does it mean for the future of medical image analysis? Machine learning has witnessed a tremendous amount of attention over the last few years. The current boom started around 2009 when so-called deep artificial neural networks began outperforming other established models on a number of important benchmarks. Deep neural networks are now the state-of-the-art machine learning models across a variety of areas, from image analysis to natural language processing, and widely deployed in academia and industry. These developments have a huge potential for medical imaging technology, medical data analysis, medical diagnostics and healthcare in general, slowly being realized. We provide a short overview of recent advances and some associated challenges in machine learning applied to medical image processing and image analysis. As this has become a very broad and fast expanding field we will not survey the entire landscape of applications, but put particular focus on deep learning in MRI. Our aim is threefold: (i) give a brief introduction to deep learning with pointers to core references; (ii) indicate how deep learning has been applied to the entire MRI processing chain, from acquisition to image retrieval, from segmentation to disease prediction; (iii) provide a starting point for people interested in experimenting and perhaps contributing to the field of machine learning for medical imaging by pointing out good educational resources, state-of-the-art open-source code, and interesting sources of data and problems related medical imaging.

arXiv:1811.09385공식 코드 · Papers with Code

On the Importance of Strong Baselines in Bayesian Deep Learning

Like all sub-fields of machine learning Bayesian Deep Learning is driven by empirical validation of its theoretical proposals. Given the many aspects of an experiment it is always possible that minor or even major experimental flaws can slip by both authors and reviewers. One of the most popular experiments used to evaluate approximate inference techniques is the regression experiment on UCI datasets. However, in this experiment, models which have been trained to convergence have often been compared with baselines trained only for a fixed number of iterations. We find that a well-established baseline, Monte Carlo dropout, when evaluated under the same experimental settings shows significant improvements. In fact, the baseline outperforms or performs competitively with methods that claimed to be superior to the very same baseline method when they were introduced. Hence, by exposing this flaw in experimental procedure, we highlight the importance of using identical experimental setups to evaluate, compare, and benchmark methods in Bayesian Deep Learning.

arXiv:1809.09441공식 코드 · Papers with Code

Temporal Relational Ranking for Stock Prediction

Stock prediction aims to predict the future trends of a stock in order to help investors to make good investment decisions. Traditional solutions for stock prediction are based on time-series models. With the recent success of deep neural networks in modeling sequential data, deep learning has become a promising choice for stock prediction. However, most existing deep learning solutions are not optimized towards the target of investment, i.e., selecting the best stock with the highest expected revenue. Specifically, they typically formulate stock prediction as a classification (to predict stock trend) or a regression problem (to predict stock price). More importantly, they largely treat the stocks as independent of each other. The valuable signal in the rich relations between stocks (or companies), such as two stocks are in the same sector and two companies have a supplier-customer relation, is not considered. In this work, we contribute a new deep learning solution, named Relational Stock Ranking (RSR), for stock prediction. Our RSR method advances existing solutions in two major aspects: 1) tailoring the deep learning models for stock ranking, and 2) capturing the stock relations in a time-sensitive manner. The key novelty of our work is the proposal of a new component in neural network modeling, named Temporal Graph Convolution, which jointly models the temporal evolution and relation network of stocks. To validate our method, we perform back-testing on the historical data of two stock markets, NYSE and NASDAQ. Extensive experiments demonstrate the superiority of our RSR method. It outperforms state-of-the-art stock prediction solutions achieving an average return ratio of 98% and 71% on NYSE and NASDAQ, respectively.

arXiv:1811.08374공식 코드 · Papers with Code

A Gray Box Interpretable Visual Debugging Approach for Deep Sequence Learning Model

Deep Learning algorithms are often used as black box type learning and they are too complex to understand. The widespread usability of Deep Learning algorithms to solve various machine learning problems demands deep and transparent understanding of the internal representation as well as decision making. Moreover, the learning models, trained on sequential data, such as audio and video data, have intricate internal reasoning process due to their complex distribution of features. Thus, a visual simulator might be helpful to trace the internal decision making mechanisms in response to adversarial input data, and it would help to debug and design appropriate deep learning models. However, interpreting the internal reasoning of deep learning model is not well studied in the literature. In this work, we have developed a visual interactive web application, namely d-DeVIS, which helps to visualize the internal reasoning of the learning model which is trained on the audio data. The proposed system allows to perceive the behavior as well as to debug the model by interactively generating adversarial audio data point. The web application of d-DeVIS is available at ddevis.herokuapp.com.

arXiv:1811.06115공식 코드 · Papers with Code

Deep Learning in the Wavelet Domain

This paper examines the possibility of, and the possible advantages to learning the filters of convolutional neural networks (CNNs) for image analysis in the wavelet domain. We are stimulated by both Mallat's scattering transform and the idea of filtering in the Fourier domain. It is important to explore new spaces in which to learn, as these may provide inherent advantages that are not available in the pixel space. However, the scattering transform is limited by its inability to learn in between scattering orders, and any Fourier domain filtering is limited by the large number of filter parameters needed to get localized filters. Instead we consider filtering in the wavelet domain with learnable filters. The wavelet space allows us to have local, smooth filters with far fewer parameters, and learnability can give us flexibility. We present a novel layer which takes CNN activations into the wavelet space, learns parameters and returns to the pixel space. This allows it to be easily dropped in to any neural network without affecting the structure. As part of this work, we show how to pass gradients through a multirate system and give preliminary results.

arXiv:1811.04132공식 코드 · Papers with Code

Reasoning over RDF Knowledge Bases using Deep Learning

Semantic Web knowledge representation standards, and in particular RDF and OWL, often come endowed with a formal semantics which is considered to be of fundamental importance for the field. Reasoning, i.e., the drawing of logical inferences from knowledge expressed in such standards, is traditionally based on logical deductive methods and algorithms which can be proven to be sound and complete and terminating, i.e. correct in a very strong sense. For various reasons, though, in particular, the scalability issues arising from the ever-increasing amounts of Semantic Web data available and the inability of deductive algorithms to deal with noise in the data, it has been argued that alternative means of reasoning should be investigated which bear high promise for high scalability and better robustness. From this perspective, deductive algorithms can be considered the gold standard regarding correctness against which alternative methods need to be tested. In this paper, we show that it is possible to train a Deep Learning system on RDF knowledge graphs, such that it is able to perform reasoning over new RDF knowledge graphs, with high precision and recall compared to the deductive gold standard.

arXiv:1811.04104공식 코드 · Papers with Code

Deep Learning Super-Diffusion in Multiplex Networks

Complex network theory has shown success in understanding the emergent and collective behavior of complex systems [1]. Many real-world complex systems were recently discovered to be more accurately modeled as multiplex networks [2-6]---in which each interaction type is mapped to its own network layer; e.g.~multi-layer transportation networks, coupled social networks, metabolic and regulatory networks, etc. A salient physical phenomena emerging from multiplexity is super-diffusion: exhibited by an accelerated diffusion admitted by the multi-layer structure as compared to any single layer. Theoretically super-diffusion was only known to be predicted using the spectral gap of the full Laplacian of a multiplex network and its interacting layers. Here we turn to machine learning which has developed techniques to recognize, classify, and characterize complex sets of data. We show that modern machine learning architectures, such as fully connected and convolutional neural networks, can classify and predict the presence of super-diffusion in multiplex networks with 94.12\% accuracy. Such predictions can be done {\it in situ}, without the need to determine spectral properties of a network.

arXiv:1811.04756공식 코드 · Papers with Code

Deep-learning the Latent Space of Light Transport

We suggest a method to directly deep-learn light transport, i. e., the mapping from a 3D geometry-illumination-material configuration to a shaded 2D image. While many previous learning methods have employed 2D convolutional neural networks applied to images, we show for the first time that light transport can be learned directly in 3D. The benefit of 3D over 2D is, that the former can also correctly capture illumination effects related to occluded and/or semi-transparent geometry. To learn 3D light transport, we represent the 3D scene as an unstructured 3D point cloud, which is later, during rendering, projected to the 2D output image. Thus, we suggest a two-stage operator comprising of a 3D network that first transforms the point cloud into a latent representation, which is later on projected to the 2D output image using a dedicated 3D-2D network in a second step. We will show that our approach results in improved quality in terms of temporal coherence while retaining most of the computational efficiency of common 2D methods. As a consequence, the proposed two stage-operator serves as a valuable extension to modern deferred shading approaches.

arXiv:1811.03691공식 코드 · Papers with Code

Can Deep Learning Outperform Modern Commercial CT Image Reconstruction Methods?

Commercial iterative reconstruction techniques on modern CT scanners target radiation dose reduction but there are lingering concerns over their impact on image appearance and low contrast detectability. Recently, machine learning, especially deep learning, has been actively investigated for CT. Here we design a novel neural network architecture for low-dose CT (LDCT) and compare it with commercial iterative reconstruction methods used for standard of care CT. While popular neural networks are trained for end-to-end mapping, driven by big data, our novel neural network is intended for end-to-process mapping so that intermediate image targets are obtained with the associated search gradients along which the final image targets are gradually reached. This learned dynamic process allows to include radiologists in the training loop to optimize the LDCT denoising workflow in a task-specific fashion with the denoising depth as a key parameter. Our progressive denoising network was trained with the Mayo LDCT Challenge Dataset, and tested on images of the chest and abdominal regions scanned on the CT scanners made by three leading CT vendors. The best deep learning based reconstructions are systematically compared to the best iterative reconstructions in a double-blinded reader study. It is found that our deep learning approach performs either comparably or favorably in terms of noise suppression and structural fidelity, and runs orders of magnitude faster than the commercial iterative CT reconstruction algorithms.

arXiv:1811.02084공식 코드 · Papers with Code

Mesh-TensorFlow: Deep Learning for Supercomputers

Batch-splitting (data-parallelism) is the dominant distributed Deep Neural Network (DNN) training strategy, due to its universal applicability and its amenability to Single-Program-Multiple-Data (SPMD) programming. However, batch-splitting suffers from problems including the inability to train very large models (due to memory constraints), high latency, and inefficiency at small batch sizes. All of these can be solved by more general distribution strategies (model-parallelism). Unfortunately, efficient model-parallel algorithms tend to be complicated to discover, describe, and to implement, particularly on large clusters. We introduce Mesh-TensorFlow, a language for specifying a general class of distributed tensor computations. Where data-parallelism can be viewed as splitting tensors and operations along the "batch" dimension, in Mesh-TensorFlow, the user can specify any tensor-dimensions to be split across any dimensions of a multi-dimensional mesh of processors. A Mesh-TensorFlow graph compiles into a SPMD program consisting of parallel operations coupled with collective communication primitives such as Allreduce. We use Mesh-TensorFlow to implement an efficient data-parallel, model-parallel version of the Transformer sequence-to-sequence model. Using TPU meshes of up to 512 cores, we train Transformer models with up to 5 billion parameters, surpassing state of the art results on WMT'14 English-to-French translation task and the one-billion-word language modeling benchmark. Mesh-Tensorflow is available at https://github.com/tensorflow/mesh .

arXiv:1811.01721공식 코드 · Papers with Code

Rethinking floating point for deep learning

Reducing hardware overhead of neural networks for faster or lower power inference and training is an active area of research. Uniform quantization using integer multiply-add has been thoroughly investigated, which requires learning many quantization parameters, fine-tuning training or other prerequisites. Little effort is made to improve floating point relative to this baseline; it remains energy inefficient, and word size reduction yields drastic loss in needed dynamic range. We improve floating point to be more energy efficient than equivalent bit width integer hardware on a 28 nm ASIC process while retaining accuracy in 8 bits with a novel hybrid log multiply/linear add, Kulisch accumulation and tapered encodings from Gustafson's posit format. With no network retraining, and drop-in replacement of all math and float32 parameters via round-to-nearest-even only, this open-sourced 8-bit log float is within 0.9% top-1 and 0.2% top-5 accuracy of the original float32 ResNet-50 CNN model on ImageNet. Unlike int8 quantization, it is still a general purpose floating point arithmetic, interpretable out-of-the-box. Our 8/38-bit log float multiply-add is synthesized and power profiled at 28 nm at 0.96x the power and 1.12x the area of 8/32-bit integer multiply-add. In 16 bits, our log float multiply-add is 0.59x the power and 0.68x the area of IEEE 754 float16 fused multiply-add, maintaining the same signficand precision and dynamic range, proving useful for training ASICs as well.

arXiv:1811.01494공식 코드 · Papers with Code

Underwater Fish Detection using Deep Learning for Water Power Applications

Clean energy from oceans and rivers is becoming a reality with the development of new technologies like tidal and instream turbines that generate electricity from naturally flowing water. These new technologies are being monitored for effects on fish and other wildlife using underwater video. Methods for automated analysis of underwater video are needed to lower the costs of analysis and improve accuracy. A deep learning model, YOLO, was trained to recognize fish in underwater video using three very different datasets recorded at real-world water power sites. Training and testing with examples from all three datasets resulted in a mean average precision (mAP) score of 0.5392. To test how well a model could generalize to new datasets, the model was trained using examples from only two of the datasets and then tested on examples from all three datasets. The resulting model could not recognize fish in the dataset that was not part of the training set. The mAP scores on the other two datasets that were included in the training set were higher than the scores achieved by the model trained on all three datasets. These results indicate that different methods are needed in order to produce a trained model that can generalize to new data sets such as those encountered in real world applications.

arXiv:1811.00908공식 코드 · Papers with Code

Single-Model Uncertainties for Deep Learning

We provide single-model estimates of aleatoric and epistemic uncertainty for deep neural networks. To estimate aleatoric uncertainty, we propose Simultaneous Quantile Regression (SQR), a loss function to learn all the conditional quantiles of a given target variable. These quantiles can be used to compute well-calibrated prediction intervals. To estimate epistemic uncertainty, we propose Orthonormal Certificates (OCs), a collection of diverse non-constant functions that map all training samples to zero. These certificates map out-of-distribution examples to non-zero values, signaling epistemic uncertainty. Our uncertainty estimators are computationally attractive, as they do not require ensembling or retraining deep models, and achieve competitive performance.

arXiv:1811.00338공식 코드 · Papers with Code

Deep Learning-Based Gait Recognition Using Smartphones in the Wild

Compared to other biometrics, gait is difficult to conceal and has the advantage of being unobtrusive. Inertial sensors, such as accelerometers and gyroscopes, are often used to capture gait dynamics. These inertial sensors are commonly integrated into smartphones and are widely used by the average person, which makes gait data convenient and inexpensive to collect. In this paper, we study gait recognition using smartphones in the wild. In contrast to traditional methods, which often require a person to walk along a specified road and/or at a normal walking speed, the proposed method collects inertial gait data under unconstrained conditions without knowing when, where, and how the user walks. To obtain good person identification and authentication performance, deep-learning techniques are presented to learn and model the gait biometrics based on walking data. Specifically, a hybrid deep neural network is proposed for robust gait feature representation, where features in the space and time domains are successively abstracted by a convolutional neural network and a recurrent neural network. In the experiments, two datasets collected by smartphones for a total of 118 subjects are used for evaluations. The experiments show that the proposed method achieves higher than 93.5\% and 93.7\% accuracies in person identification and authentication, respectively.