PyTorchKR Papers

최신 논문

총 8,922편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:1911.12461

Two-Stage Learning for Uplink Channel Estimation in One-Bit Massive MIMO

We develop a two-stage deep learning pipeline architecture to estimate the uplink massive MIMO channel with one-bit ADCs. This deep learning pipeline is composed of two separate generative deep learning models. The first one is a supervised learning model and designed to compensate for the quantization loss. The second one is an unsupervised learning model and optimized for denoising. Our results show that the proposed deep learning-based channel estimator can significantly outperform other state-of-the-art channel estimators for one-bit quantized massive MIMO systems. In particular, our design provides 5-10 dB gain in channel estimation error. Furthermore, it requires a reasonable amount of pilots, on the order of 20 per coherence time interval.

arXiv:1911.04798

MRI denoising using Deep Learning and Non-local averaging

This paper proposes a novel method for automatic MRI denoising that exploits last advances in deep learning feature regression and self-similarity properties of the MR images. The proposed method is a two-stage approach. In the first stage, an overcomplete patch-based convolutional neural network blindly removes the noise without specific estimation of the local noise variance to produce a preliminary estimation of the noise-free image. The second stage uses this preliminary denoised image as a guide image within a rotationally invariant non-local means filter to robustly denoise the original noisy image. The proposed approach has been compared with related state-of-the-art methods and showed competitive results in all the studied cases while being much faster than comparable filters. We present a denoising method that can be blindly applied to any type of MR image since it can automatically deal with both stationary and spatially varying noise patterns.

arXiv:1910.13906

Probabilistic performance validation of deep learning-based robust NMPC controllers

Solving nonlinear model predictive control problems in real time is still an important challenge despite of recent advances in computing hardware, optimization algorithms and tailored implementations. This challenge is even greater when uncertainty is present due to disturbances, unknown parameters or measurement and estimation errors. To enable the application of advanced control schemes to fast systems and on low-cost embedded hardware, we propose to approximate a robust nonlinear model controller using deep learning and to verify its quality using probabilistic validation techniques. We propose a probabilistic validation technique based on finite families, combined with the idea of generalized maximum and constraint backoff to enable statistically valid conclusions related to general performance indicators. The potential of the proposed approach is demonstrated with simulation results of an uncertain nonlinear system.

arXiv:1910.11402

Deep learning-based beam alignment in mmWave vehicular networks

Millimeter wave channels exhibit structure that allows beam alignment with fewer channel measurements than exhaustive beam search. From a compressed sensing (CS) perspective, the received channel measurements are usually obtained by multiplying a CS matrix with a sparse representation of the channel matrix. Due to the constraints imposed by analog processing, designing CS matrices that efficiently exploit the channel structure is, however, challenging. In this paper, we propose an end-to-end deep learning technique to design a structured CS matrix that is well suited to the underlying channel distribution, leveraging both sparsity and the particular spatial structure that appears in vehicular channels. The channel measurements acquired with the designed CS matrix are then used to predict the best beam for link configuration. Simulation results for vehicular communication channels indicate that our deep learning-based approach achieves better beam alignment than standard CS techniques that use the random phase shift-based design.

arXiv:1910.09136

Deep-Learning-Aided Detection for Reconfigurable Intelligent Surfaces

This paper presents a deep learning (DL) approach for estimating and detecting symbols in signals transmitted through reconfigurable intelligent surfaces (RIS). The proposed network utilizes fully connected layers to estimate channels and phase angles from a reflected signal received through an RIS. Because the proposed network can estimate and detect symbols without any pilot signaling, this method reduces the overhead required for transmission. The improvements achieved by this method are quantified in terms of the bit-error rate, outperforming traditional detectors.

arXiv:1910.06659

Ballistocardiogram artifact reduction in simultaneous EEG-fMRI using deep learning

Objective: The concurrent recording of electroencephalography (EEG) and functional magnetic resonance imaging (fMRI) is a technique that has received much attention due to its potential for combined high temporal and spatial resolution. However, the ballistocardiogram (BCG), a large-amplitude artifact caused by cardiac induced movement contaminates the EEG during EEG-fMRI recordings. Removal of BCG in software has generally made use of linear decompositions of the corrupted EEG. This is not ideal as the BCG signal is non-stationary and propagates in a manner which is non-linearly dependent on the electrocardiogram (ECG). In this paper, we present a novel method for BCG artifact suppression using recurrent neural networks (RNNs). Methods: EEG signals were recovered by training RNNs on the nonlinear mappings between ECG and the BCG corrupted EEG. We evaluated our model's performance against the commonly used Optimal Basis Set (OBS) method at the level of individual subjects, and investigated generalization across subjects. Results: We show that our algorithm can generate larger average power reduction of the BCG at critical frequencies, while simultaneously improving task relevant EEG based classification. Conclusion: The presented deep learning architecture can be used to reduce BCG related artifacts in EEG-fMRI recordings. Significance: We present a deep learning approach that can be used to suppress the BCG artifact in EEG-fMRI without the use of additional hardware. This method may have scope to be combined with current hardware methods, operate in real-time and be used for direct modeling of the BCG.

arXiv:1909.10342

Adaptive Ultrasound Beamforming using Deep Learning

Biomedical imaging is unequivocally dependent on the ability to reconstruct interpretable and high-quality images from acquired sensor data. This reconstruction process is pivotal across many applications, spanning from magnetic resonance imaging to ultrasound imaging. While advanced data-adaptive reconstruction methods can recover much higher image quality than traditional approaches, their implementation often poses a high computational burden. In ultrasound imaging, this burden is significant, especially when striving for low-cost systems, and has motivated the development of high-resolution and high-contrast adaptive beamforming methods. Here we show that deep neural networks that adopt the algorithmic structure and constraints of adaptive signal processing techniques can efficiently learn to perform fast high-quality ultrasound beamforming using very little training data. We apply our technique to two distinct ultrasound acquisition strategies (plane wave, and synthetic aperture), and demonstrate that high image quality can be maintained when measuring at low data-rates, using undersampled array designs. Beyond biomedical imaging, we expect that the proposed deep~learning based adaptive processing framework can benefit a variety of array and signal processing applications, in particular when data-efficiency and robustness are of importance.

arXiv:1909.07784

MathDL: Mathematical deep learning for D3R Grand Challenge 4

We present the performances of our mathematical deep learning (MathDL) models for D3R Grand Challenge 4 (GC4). This challenge involves pose prediction, affinity ranking, and free energy estimation for beta secretase 1 (BACE) as well as affinity ranking and free energy estimation for Cathepsin S (CatS). We have developed advanced mathematics, namely differential geometry, algebraic graph, and/or algebraic topology, to accurately and efficiently encode high dimensional physical/chemical interactions into scalable low-dimensional rotational and translational invariant representations. These representations are integrated with deep learning models, such as generative adversarial networks (GAN) and convolutional neural networks (CNN) for pose prediction and energy evaluation, respectively. Overall, our MathDL models achieved the top place in pose prediction for BACE ligands in Stage 1a. Moreover, our submissions obtained the highest Spearman correlation coefficient on the affinity ranking of 460 CatS compounds, and the smallest centered root mean square error on the free energy set of 39 CatS molecules. It is worthy to mention that our method for docking pose predictions has significantly improved from our previous ones.

arXiv:1909.06031

Deep Learning for Cooperative Radio Signal Classification

Radio signal classification has a very wide range of applications in cognitive radio networks and electromagnetic spectrum monitoring. In this article, we consider scenarios where multiple nodes in the network participate in cooperative classification. We propose cooperative radio signal classification methods based on deep learning for decision fusion, signal fusion and feature fusion, respectively. We analyze the performance of these methods through simulation experiments. We conclude the article with a discussion of research challenges and open problems.

arXiv:1909.06016

Enhancement of seismic imaging: An innovative deep learning approach

Enhancing the frequency bandwidth of the seismic data is always the pursuance at the geophysical community. High resolution of seismic data provides the key resource to extract detailed stratigraphic knowledge. Here, a novel approach, based on deep learning model, is introduced by extracting reflections from well log data to broaden spectrum bandwidth of seismic data through boosting low and high frequencies. The corresponding improvement is observed from the enhancement of resolution of seismic data as well as elimination of sidelobe artifacts from seismic wavelets. During the training stage of deep learning model, geo-spatial information by taking consideration of multiple wells simultaneously is fully guaranteed, which assures that laterally and vertically geological information are constrained by and accurate away from the well controls during the inversion procedure. Extensive experiments prove that the enhanced seismic data is consistent with well log information, and honors rock property relationships defined from the wells at given locations. Uncertainty analysis could also be quantitatively assessed to determine the possibilities of a range of seismic responses by leveraging the outputs from the proposed approach.

arXiv:1908.11013

Deep Learning based Channel Estimation Algorithm over Time Selective Fading Channels

The research about deep learning application for physical layer has been received much attention in recent years. In this paper, we propose a Deep Learning (DL) based channel estimator under time varying Rayleigh fading channel. We build up, train and test the channel estimator using Neural Network (NN). The proposed DL-based estimator can dynamically track the channel status without any prior knowledge about the channel model and statistic characteristics. The simulation results show the proposed NN estimator has better Mean Square Error (MSE) performance compared with the traditional algorithms and some other DL-based architectures. Furthermore, the proposed DL-based estimator also shows its robustness with the different pilot densities.

arXiv:1908.09314

Surround Inhibition Mechanism by Deep Learning

In the sensation of tones, visions and other stimuli, the "surround inhibition mechanism" (or "lateral inhibition mechanism") is crucial. The mechanism enhances the signals of the strongest tone, color and other stimuli, by reducing and inhibiting the surrounding signals, since the latter signals are less important. This surround inhibition mechanism is well studied in the physiology of sensor systems. The neural network with two hidden layers in addition to input and output layers is constructed; having 60 neurons (units) in each of the four layers. The label (correct answer) is prepared from an input signal by applying seven times operations of the "Hartline mechanism", that is, by sending inhibitory signals from the neighboring neurons and amplifying all the signals afterwards. The implication obtained by the deep learning of this neural network is compared with the standard physiological understanding of the surround inhibition mechanism.

arXiv:1908.07488

LIDAR Data for Deep Learning-Based mmWave Beam-Selection

Millimeter wave communication systems can leverage information from sensors to reduce the overhead associated with link configuration. LIDAR (light detection and ranging) is one sensor widely used in autonomous driving for high resolution mapping and positioning. This paper shows how LIDAR data can be used for line-of-sight detection and to reduce the overhead in millimeter wave beam-selection. In the proposed distributed architecture, the base station broadcasts its position. The connected vehicle leverages its LIDAR data to suggest a set of beams selected via a deep convolutional neural network. Co-simulation of communications and LIDAR in a vehicle-to-infrastructure (V2I) scenario confirm that LIDAR can help configuring mmWave V2I links.

arXiv:1907.11953

Automated Mammogram Analysis with a Deep Learning Pipeline

Current deep learning based detection models tackle detection and segmentation tasks by casting them to pixel or patch-wise classification. To automate the initial mass lesion detection and segmentation on the whole mammographic images and avoid the computational redundancy of patch-based and sliding window approaches, the conditional generative adversarial network (cGAN) was used in this study. Subsequently, feeding the detected regions to the trained densely connected network (DenseNet), the binary classification of benign versus malignant was predicted. We used a combination of publicly available mammographic data repositories to train the pipeline, while evaluating the model's robustness toward our clinically collected repository, which was unseen to the pipeline.

arXiv:1906.07319공식 코드 · Papers with Code

Deep Xi as a Front-End for Robust Automatic Speech Recognition

Current front-ends for robust automatic speech recognition(ASR) include masking- and mapping-based deep learning approaches to speech enhancement. A recently proposed deep learning approach toa prioriSNR estimation, called DeepXi, was able to produce enhanced speech at a higher quality and intelligibility than current masking- and mapping-based approaches. Motivated by this, we investigate Deep Xi as a front-end for robust ASR. Deep Xi is evaluated using real-world non-stationary and coloured noise sources at multiple SNR levels. Our experimental investigation shows that DeepXi as a front-end is able to produce a lower word error rate than recent masking- and mapping-based deep learning front-ends. The results presented in this work show that Deep Xi is a viable front-end, and is able to significantly increase the robustness of an ASR system. Availability: Deep Xi is available at:https://github.com/anicolson/DeepXi

arXiv:1906.06225

Control-Aware Scheduling for Low Latency Wireless Systems with Deep Learning

We consider the problem of scheduling transmissions over low-latency wireless communication links to control various control systems. Low-latency requirements are critical in developing wireless technology for industrial control and Tactile Internet, but are inherently challenging to meet while also maintaining reliable performance. An alternative to ultra reliable low latency communications is a framework in which reliability is adapted to control system demands. We formulate the control-aware scheduling problem as a constrained statistical optimization problem in which the optimal scheduler is a function of current control and channel states. The scheduler is parameterized with a deep neural network, and the constrained problem is solved using techniques from primal-dual learning, which have a necessary model-free property in that they do not require explicit knowledge of channels models, performance metrics, or system dynamics to execute. The resulting control-aware deep scheduler is evaluated in empirical simulations and strong performance is shown relative to other model-free heuristic scheduling methods.

arXiv:1905.10468

A Deep Learning Wireless Transceiver with Fully Learned Modulation and Synchronization

In this paper, we present a deep learning based wireless transceiver. We describe in detail the corresponding artificial neural network architecture, the training process, and report on excessive over-the-air measurement results. We employ the end-to-end training approach with an autoencoder model that includes a channel model in the middle layers as previously proposed in the literature. In contrast to other state-of-the-art results, our architecture supports learning time synchronization without any manually designed signal processing operations. Moreover, the neural transceiver has been tested over the air with an implementation in software defined radio. Our experimental results for the implemented single antenna system demonstrate a raw bit-rate of 0.5 million bits per second. This exceeds results from comparable systems presented in the literature and suggests the feasibility of high throughput deep learning transceivers.

arXiv:1904.09346

Deep Learning-Based Channel Estimation for High-Dimensional Signals

We propose a novel deep learning-based channel estimation technique for high-dimensional communication signals that does not require any training. Our method is broadly applicable to channel estimation for multicarrier signals with any number of antennas, and has low enough complexity to be used in a mobile station. The proposed deep channel estimator can outperform LS estimation with nearly the same complexity, and approach MMSE estimation performance to within 1 dB without knowing the second order statistics. The only complexity increase with respect to LS estimator lies in fitting the parameters of a deep neural network (DNN) periodically on the order of the channel coherence time. We empirically show that the main benefit of this method accrues from the ability of this specially designed DNN to exploit correlations in the time-frequency grid. The proposed estimator can also reduce the number of pilot tones needed in an OFDM time-frequency grid, e.g. in an LTE scenario by 98% (68%) when the channel coherence time interval is 73ms (4.5ms).

arXiv:1904.05921

Deep-learning based numerical BSDE method for barrier options

As is known, an option price is a solution to a certain partial differential equation (PDE) with terminal conditions (payoff functions). There is a close association between the solution of PDE and the solution of a backward stochastic differential equation (BSDE). We can either solve the PDE to obtain option prices or solve its associated BSDE. Recently a deep learning technique has been applied to solve option prices using the BSDE approach. In this approach, deep learning is used to learn some deterministic functions, which are used in solving the BSDE with terminal conditions. In this paper, we extend the deep-learning technique to solve a PDE with both terminal and boundary conditions. In particular, we will employ the technique to solve barrier options using Brownian motion bridges.

arXiv:1904.00745

Deep Learning in Asset Pricing

We use deep neural networks to estimate an asset pricing model for individual stock returns that takes advantage of the vast amount of conditioning information, while keeping a fully flexible form and accounting for time-variation. The key innovations are to use the fundamental no-arbitrage condition as criterion function, to construct the most informative test assets with an adversarial approach and to extract the states of the economy from many macroeconomic time series. Our asset pricing model outperforms out-of-sample all benchmark approaches in terms of Sharpe ratio, explained variation and pricing errors and identifies the key factors that drive asset prices.