PyTorchKR Papers

최신 논문

총 3,337편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2006.02000

MultiXNet: Multiclass Multistage Multimodal Motion Prediction

One of the critical pieces of the self-driving puzzle is understanding the surroundings of a self-driving vehicle (SDV) and predicting how these surroundings will change in the near future. To address this task we propose MultiXNet, an end-to-end approach for detection and motion prediction based directly on lidar sensor data. This approach builds on prior work by handling multiple classes of traffic actors, adding a jointly trained second-stage trajectory refinement step, and producing a multimodal probability distribution over future actor motion that includes both multiple discrete traffic behaviors and calibrated continuous position uncertainties. The method was evaluated on large-scale, real-world data collected by a fleet of SDVs in several cities, with the results indicating that it outperforms existing state-of-the-art approaches.

arXiv:2005.12992

Evaluating Information Retrieval Systems for Kids

Evaluation of information retrieval systems (IRS) is a prominent topic among information retrieval researchers--mainly directed at a general population. Children require unique IRS and by extension different ways to evaluate these systems, but as a large population that use IRS have largely been ignored on the evaluation front. In this position paper, we explore many perspectives that must be considered when evaluating IRS; we specially discuss problems faced by researchers who work with children IRS, including lack of evaluation frameworks, limitations of data, and lack of user judgment understanding.

arXiv:2005.13458공식 코드 · Papers with Code

Fast Risk Assessment for Autonomous Vehicles Using Learned Models of Agent Futures

This paper presents fast non-sampling based methods to assess the risk of trajectories for autonomous vehicles when probabilistic predictions of other agents' futures are generated by deep neural networks (DNNs). The presented methods address a wide range of representations for uncertain predictions including both Gaussian and non-Gaussian mixture models for predictions of both agent positions and controls. We show that the problem of risk assessment when Gaussian mixture models (GMMs) of agent positions are learned can be solved rapidly to arbitrary levels of accuracy with existing numerical methods. To address the problem of risk assessment for non-Gaussian mixture models of agent position, we propose finding upper bounds on risk using Chebyshev's Inequality and sums-of-squares (SOS) programming; they are both of interest as the former is much faster while the latter can be arbitrarily tight. These approaches only require statistical moments of agent positions to determine upper bounds on risk. To perform risk assessment when models are learned for agent controls as opposed to positions, we develop TreeRing, an algorithm analogous to tree search over the ring of polynomials that can be used to exactly propagate moments of control distributions into position distributions through nonlinear dynamics. The presented methods are demonstrated on realistic predictions from DNNs trained on the Argoverse and CARLA datasets and are shown to be effective for rapidly assessing the probability of low probability events.

arXiv:2005.10904

Solving a steady-state PDE using spiking networks and neuromorphic hardware

The widely parallel, spiking neural networks of neuromorphic processors can enable computationally powerful formulations. While recent interest has focused on primarily machine learning tasks, the space of appropriate applications is wide and continually expanding. Here, we leverage the parallel and event-driven structure to solve a steady state heat equation using a random walk method. The random walk can be executed fully within a spiking neural network using stochastic neuron behavior, and we provide results from both IBM TrueNorth and Intel Loihi implementations. Additionally, we position this algorithm as a potential scalable benchmark for neuromorphic systems.

논문 ID: document/8853556

Implementation of a WiFi-based indoor location system on a mobile device for a university area

This paper describes the implementation of an indoor location system on a mobile phone for the Faculty of Electrical, Electronic and Telecomunications at the Universidad Nacional de Ingeniería, Peru. The proposed system makes use of WiFi signals and the Bayes filter in order to predict the user location. The principal advantage of the proposed method in comparison with similar methodologies is that our system does not need additional hardware or special equipments. As the experimental results show, the system achieves 92.31% for position accuracy and works in real-time. In addition, the proposed methodology could be used for similar location tasks.

arXiv:2005.11106

On the suitability of generalized regression neural networks for GNSS position time series prediction for geodetic applications in geodesy and geophysics

In this paper, the generalized regression neural network is used to predict the GNSS position time series. Using the IGS 24-hour final solution data for Bad Hamburg permanent GNSS station in Germany, it is shown that the larger the training of the network, the higher the accuracy is, regardless of the time span of the time series. In order to analyze the performance of the neural network in various conditions, 14 permanent stations are used in different countries, namely, Spain, France, Romania, Poland, Russian Federation, United Kingdom, Czech Republic, Sweden, Ukraine, Italy, Finland, Slovak Republic, Cyprus, and Greece. The performance analysis is divided into two parts, continuous data-without gaps-and discontinuous ones-having intervals of gaps with no data available. Three measure of error are presented, namely, symmetric mean absolute percentage error, standard deviation, and mean of absolute errors. It is shown that for discontinuous data the position can be predicted with an accuracy of up to 6 centimeters, while the continuous data positions present a higher prediction accuracy, as high as 3 centimeters. In order to compare the results of this machine learning algorithm with the traditional statistical approaches, the Theta method is used, which is well-established for high-accuracy time series prediction. The comparison shows that the generalized regression neural network machine learning algorithm presents better accuracy than the Theta method, possibly up to 250 times. In addition, it is approximately 4.6 times faster.

arXiv:2005.11035공식 코드 · Papers with Code

Position-based Scaled Gradient for Model Quantization and Pruning

We propose the position-based scaled gradient (PSG) that scales the gradient depending on the position of a weight vector to make it more compression-friendly. First, we theoretically show that applying PSG to the standard gradient descent (GD), which is called PSGD, is equivalent to the GD in the warped weight space, a space made by warping the original weight space via an appropriately designed invertible function. Second, we empirically show that PSG acting as a regularizer to a weight vector is favorable for model compression domains such as quantization and pruning. PSG reduces the gap between the weight distributions of a full-precision model and its compressed counterpart. This enables the versatile deployment of a model either as an uncompressed mode or as a compressed mode depending on the availability of resources. The experimental results on CIFAR-10/100 and ImageNet datasets show the effectiveness of the proposed PSG in both domains of pruning and quantization even for extremely low bits. The code is released in Github.

arXiv:2005.09940

Relative Positional Encoding for Speech Recognition and Direct Translation

Transformer models are powerful sequence-to-sequence architectures that are capable of directly mapping speech inputs to transcriptions or translations. However, the mechanism for modeling positions in this model was tailored for text modeling, and thus is less ideal for acoustic inputs. In this work, we adapt the relative position encoding scheme to the Speech Transformer, where the key addition is relative distance between input states in the self-attention network. As a result, the network can better adapt to the variable distributions present in speech data. Our experiments show that our resulting model achieves the best recognition result on the Switchboard benchmark in the non-augmentation condition, and the best published result in the MuST-C speech translation benchmark. We also show that this model is able to better utilize synthetic data than the Transformer, and adapts better to variable sentence segmentation quality for speech translation.

arXiv:1612.02261

Sparse Geometric Representation Through Local Shape Probing

We propose a new shape analysis approach based on the non-local analysis of local shape variations. Our method relies on a novel description of shape variations, called Local Probing Field (LPF), which describes how a local probing operator transforms a pattern onto the shape. By carefully optimizing the position and orientation of each descriptor, we are able to capture shape similarities and gather them into a geometrically relevant dictionary over which the shape decomposes sparsely. This new representation permits to handle shapes with mixed intrinsic dimensionality (e.g. shapes containing both surfaces and curves) and to encode various shape features such as boundaries. Our shape representation has several potential applications; here we demonstrate its efficiency for shape resampling and point set denoising for both synthetic and real data

arXiv:2005.08245

Dampen the Stop-and-Go Traffic with Connected and Automated Vehicles -- A Deep Reinforcement Learning Approach

Stop-and-go traffic poses many challenges to tranportation system, but its formation and mechanism are still under exploration.however, it has been proved that by introducing Connected Automated Vehicles(CAVs) with carefully designed controllers one could dampen the stop-and-go waves in the vehicle fleet. Instead of using analytical model, this study adopts reinforcement learning to control the behavior of CAV and put a single CAV at the 2nd position of a vehicle fleet with the purpose to dampen the speed oscillation from the fleet leader and help following human drivers adopt more smooth driving behavior. The result show that our controller could decrease the spped oscillation of the CAV by 54% and 8%-28% for those following human-driven vehicles. Significant fuel consumption savings are also observed. Additionally, the result suggest that CAVs may act as a traffic stabilizer if they choose to behave slightly altruistically.

arXiv:2005.08919

Modeling extra-deep electromagnetic logs using a deep neural network

Modern geosteering is heavily dependent on real-time interpretation of deep electromagnetic (EM) measurements. We present a methodology to construct a deep neural network (DNN) model trained to reproduce a full set of extra-deep EM logs consisting of 22 measurements per logging position. The model is trained in a 1D layered environment consisting of up to seven layers with different resistivity values. A commercial simulator provided by a tool vendor is used to generate a training dataset. The dataset size is limited because the simulator provided by the vendor is optimized for sequential execution. Therefore, we design a training dataset that embraces the geological rules and geosteering specifics supported by the forward model. We use this dataset to produce an EM simulator based on a DNN without access to the proprietary information about the EM tool configuration or the original simulator source code. Despite employing a relatively small training set size, the resulting DNN forward model is quite accurate for the considered examples: a multi-layer synthetic case and a section of a published historical operation from the Goliat Field. The observed average evaluation time of 0.15 ms per logging position makes it also suitable for future use as part of evaluation-hungry statistical and/or Monte-Carlo inversion algorithms within geosteering workflows.

arXiv:2005.07421

Spelling Error Correction with Soft-Masked BERT

Spelling error correction is an important yet challenging task because a satisfactory solution of it essentially needs human-level language understanding ability. Without loss of generality we consider Chinese spelling error correction (CSC) in this paper. A state-of-the-art method for the task selects a character from a list of candidates for correction (including non-correction) at each position of the sentence on the basis of BERT, the language representation model. The accuracy of the method can be sub-optimal, however, because BERT does not have sufficient capability to detect whether there is an error at each position, apparently due to the way of pre-training it using mask language modeling. In this work, we propose a novel neural architecture to address the aforementioned issue, which consists of a network for error detection and a network for error correction based on BERT, with the former being connected to the latter with what we call soft-masking technique. Our method of using `Soft-Masked BERT' is general, and it may be employed in other language detection-correction problems. Experimental results on two datasets demonstrate that the performance of our proposed method is significantly better than the baselines including the one solely based on BERT.

arXiv:2005.04513

Intelligent GPS Spoofing Attack Detection in Power Grids

The GPS is vulnerable to GPS spoofing attack (GSA), which leads to disorder in time and position results of the GPS receiver. In power grids, phasor measurement units (PMUs) use GPS to build time-tagged measurements, so they are susceptible to this attack. As a result of this attack, sampling time and phase angle of the PMU measurements change. In this paper, a neural network GPS spoofing detection (NNGSD) with employing PMU data from the dynamic power system is presented to detect GSAs. Numerical results in different conditions show the real-time performance of the proposed detection method.

arXiv:2005.04954

Propagation Graph Estimation from Individual's Time Series of Observed States

Various things propagate through the medium of individuals. Some individuals follow the others and take the states similar to their states a small number of time steps later. In this paper, we study the problem of estimating the state propagation order of individuals from the real-valued state sequences of all the individuals. We propose a method to estimate the propagation direction between individuals by the sum of the time delay of one individual's state positions from the other individual's matched state position averaged over the minimum cost alignments and show how to calculate it efficiently. The propagation order estimated by our proposed method is demonstrated to be significantly more accurate than that by a baseline method for our synthetic datasets, and also to be consistent with visually recognizable propagation orders for the dataset of Japanese stock price time series and biological cell firing state sequences.

arXiv:2005.04848

Non-iterative Simultaneous Rigid Registration Method for Serial Sections of Biological Tissue

In this paper, we propose a novel non-iterative algorithm to simultaneously estimate optimal rigid transformation for serial section images, which is a key component in volume reconstruction of serial sections of biological tissue. In order to avoid error accumulation and propagation caused by current algorithms, we add extra condition that the position of the first and the last section images should remain unchanged. This constrained simultaneous registration problem has not been solved before. Our algorithm method is non-iterative, it can simultaneously compute rigid transformation for a large number of serial section images in a short time. We prove that our algorithm gets optimal solution under ideal condition. And we test our algorithm with synthetic data and real data to verify our algorithm's effectiveness.

arXiv:2005.04755공식 코드 · Papers with Code

BayesRace: Learning to race autonomously using prior experience

Autonomous race cars require perception, estimation, planning, and control modules which work together asynchronously while driving at the limit of a vehicle's handling capability. A fundamental challenge encountered in designing these software components lies in predicting the vehicle's future state (e.g. position, orientation, and speed) with high accuracy. The root cause is the difficulty in identifying vehicle model parameters that capture the effects of lateral tire slip. We present a model-based planning and control framework for autonomous racing that significantly reduces the effort required in system identification and control design. Our approach alleviates the gap induced by simulation-based controller design by learning from on-board sensor measurements. A major focus of this work is empirical, thus, we demonstrate our contributions by experiments on validated 1:43 and 1:10 scale autonomous racing simulations.

arXiv:2005.04209

BCI-Controlled Hands-Free Wheelchair Navigation with Obstacle Avoidance

Brain-Computer interfaces (BCI) are widely used in reading brain signals and converting them into real-world motion. However, the signals produced from the BCI are noisy and hard to analyze. This paper looks specifically towards combining the BCI's latest technology with ultrasonic sensors to provide a hands-free wheelchair that can efficiently navigate through crowded environments. This combination provides safety and obstacle avoidance features necessary for the BCI Navigation system to gain more confidence and operate the wheelchair at a relatively higher velocity. A population of six human subjects tested the BCI-controller and obstacle avoidance features. Subjects were able to mentally control the destination of the wheelchair, by moving the target from the starting position to a predefined position, in an average of 287.12 seconds and a standard deviation of 48.63 seconds after 10 minutes of training. The wheelchair successfully avoided all obstacles placed by the subjects during the test.

arXiv:2005.03776

Mapping Natural Language Instructions to Mobile UI Action Sequences

We present a new problem: grounding natural language instructions to mobile user interface actions, and create three new datasets for it. For full task evaluation, we create PIXELHELP, a corpus that pairs English instructions with actions performed by people on a mobile UI emulator. To scale training, we decouple the language and action data by (a) annotating action phrase spans in HowTo instructions and (b) synthesizing grounded descriptions of actions for mobile user interfaces. We use a Transformer to extract action phrase tuples from long-range natural language instructions. A grounding Transformer then contextually represents UI objects using both their content and screen position and connects them to object descriptions. Given a starting screen and instruction, our model achieves 70.59% accuracy on predicting complete ground-truth action sequences in PIXELHELP.

arXiv:2005.03482

AN-GCN: An Anonymous Graph Convolutional Network Defense Against Edge-Perturbing Attack

Recent studies have revealed the vulnerability of graph convolutional networks (GCNs) to edge-perturbing attacks, such as maliciously inserting or deleting graph edges. However, a theoretical proof of such vulnerability remains a big challenge, and effective defense schemes are still open issues. In this paper, we first generalize the formulation of edge-perturbing attacks and strictly prove the vulnerability of GCNs to such attacks in node classification tasks. Following this, an anonymous graph convolutional network, named AN-GCN, is proposed to counter against edge-perturbing attacks. Specifically, we present a node localization theorem to demonstrate how the GCN locates nodes during its training phase. In addition, we design a staggered Gaussian noise based node position generator, and devise a spectral graph convolution based discriminator in detecting the generated node positions. Further, we give the optimization of the above generator and discriminator. AN-GCN can classify nodes without taking their position as input. It is demonstrated that the AN-GCN is secure against edge-perturbing attacks in node classification tasks, as AN-GCN classifies nodes without the edge information and thus makes it impossible for attackers to perturb edges anymore. Extensive evaluations demonstrated the effectiveness of the general edge-perturbing attack model in manipulating the classification results of the target nodes. More importantly, the proposed AN-GCN can achieve 82.7% in node classification accuracy without the edge-reading permission, which outperforms the state-of-the-art GCN.

arXiv:2005.03345

Regression Forest-Based Atlas Localization and Direction Specific Atlas Generation for Pancreas Segmentation

This paper proposes a fully automated atlas-based pancreas segmentation method from CT volumes utilizing atlas localization by regression forest and atlas generation using blood vessel information. Previous probabilistic atlas-based pancreas segmentation methods cannot deal with spatial variations that are commonly found in the pancreas well. Also, shape variations are not represented by an averaged atlas. We propose a fully automated pancreas segmentation method that deals with two types of variations mentioned above. The position and size of the pancreas is estimated using a regression forest technique. After localization, a patient-specific probabilistic atlas is generated based on a new image similarity that reflects the blood vessel position and direction information around the pancreas. We segment it using the EM algorithm with the atlas as prior followed by the graph-cut. In evaluation results using 147 CT volumes, the Jaccard index and the Dice overlap of the proposed method were 62.1% and 75.1%, respectively. Although we automated all of the segmentation processes, segmentation results were superior to the other state-of-the-art methods in the Dice overlap.