PyTorchKR Papers

최신 논문

총 3,383편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:1905.07005

How do neural networks see depth in single images?

Deep neural networks have lead to a breakthrough in depth estimation from single images. Recent work often focuses on the accuracy of the depth map, where an evaluation on a publicly available test set such as the KITTI vision benchmark is often the main result of the article. While such an evaluation shows how well neural networks can estimate depth, it does not show how they do this. To the best of our knowledge, no work currently exists that analyzes what these networks have learned. In this work we take the MonoDepth network by Godard et al. and investigate what visual cues it exploits for depth estimation. We find that the network ignores the apparent size of known obstacles in favor of their vertical position in the image. Using the vertical position requires the camera pose to be known; however we find that MonoDepth only partially corrects for changes in camera pitch and roll and that these influence the estimated depth towards obstacles. We further show that MonoDepth's use of the vertical image position allows it to estimate the distance towards arbitrary obstacles, even those not appearing in the training set, but that it requires a strong edge at the ground contact point of the object to do so. In future work we will investigate whether these observations also apply to other neural networks for monocular depth estimation.

arXiv:1905.06048공식 코드 · Papers with Code

Materials property prediction using symmetry-labeled graphs as atomic-position independent descriptors

Computational materials screening studies require fast calculation of the properties of thousands of materials. The calculations are often performed with Density Functional Theory (DFT), but the necessary computer time sets limitations for the investigated material space. Therefore, the development of machine learning models for prediction of DFT calculated properties are currently of interest. A particular challenge for \emph{new} materials is that the atomic positions are generally not known. We present a machine learning model for the prediction of DFT-calculated formation energies based on Voronoi quotient graphs and local symmetry classification without the need for detailed information about atomic positions. The model is implemented as a message passing neural network and tested on the Open Quantum Materials Database (OQMD) and the Materials Project database. The test mean absolute error is 20 meV on the OQMD database and 40 meV on Materials Project Database. The possibilities for prediction in a realistic computational screening setting is investigated on a dataset of 5976 ABSe$_3$ selenides with very limited overlap with the OQMD training set. Pretraining on OQMD and subsequent training on 100 selenides result in a mean absolute error below 0.1 eV for the formation energy of the selenides.

arXiv:1905.02106

Localizing Adverts in Outdoor Scenes

Online videos have witnessed an unprecedented growth over the last decade, owing to wide range of content creation. This provides the advertisement and marketing agencies plethora of opportunities for targeted advertisements. Such techniques involve replacing an existing advertisement in a video frame, with a new advertisement. However, such post-processing of online videos is mostly done manually by video editors. This is cumbersome and time-consuming. In this paper, we propose DeepAds -- a deep neural network, based on the simple encoder-decoder architecture, that can accurately localize the position of an advert in a video frame. Our approach of localizing billboards in outdoor scenes using neural nets, is the first of its kind, and achieves the best performance. We benchmark our proposed method with other semantic segmentation algorithms, on a public dataset of outdoor scenes with manually annotated billboard binary maps.

arXiv:1905.01058

Semantic Segmentation of Video Sequences with Convolutional LSTMs

Most of the semantic segmentation approaches have been developed for single image segmentation, and hence, video sequences are currently segmented by processing each frame of the video sequence separately. The disadvantage of this is that temporal image information is not considered, which improves the performance of the segmentation approach. One possibility to include temporal information is to use recurrent neural networks. However, there are only a few approaches using recurrent networks for video segmentation so far. These approaches extend the encoder-decoder network architecture of well-known segmentation approaches and place convolutional LSTM layers between encoder and decoder. However, in this paper it is shown that this position is not optimal, and that other positions in the network exhibit better performance. Nowadays, state-of-the-art segmentation approaches rarely use the classical encoder-decoder structure, but use multi-branch architectures. These architectures are more complex, and hence, it is more difficult to place the recurrent units at a proper position. In this work, the multi-branch architectures are extended by convolutional LSTM layers at different positions and evaluated on two different datasets in order to find the best one. It turned out that the proposed approach outperforms the pure CNN-based approach for up to 1.6 percent.

arXiv:1905.00753

Galaxy Learning -- A Position Paper

The recent rapid development of artificial intelligence (AI, mainly driven by machine learning research, especially deep learning) has achieved phenomenal success in various applications. However, to further apply AI technologies in real-world context, several significant issues regarding the AI ecosystem should be addressed. We identify the main issues as data privacy, ownership, and exchange, which are difficult to be solved with the current centralized paradigm of machine learning training methodology. As a result, we propose a novel model training paradigm based on blockchain, named Galaxy Learning, which aims to train a model with distributed data and to reserve the data ownership for their owners. In this new paradigm, encrypted models are moved around instead, and are federated once trained. Model training, as well as the communication, is achieved with blockchain and its smart contracts. Pricing of training data is determined by its contribution, and therefore it is not about the exchange of data ownership. In this position paper, we describe the motivation, paradigm, design, and challenges as well as opportunities of Galaxy Learning.

arXiv:1904.11761

Factored Contextual Policy Search with Bayesian Optimization

Scarce data is a major challenge to scaling robot learning to truly complex tasks, as we need to generalize locally learned policies over different task contexts. Contextual policy search offers data-efficient learning and generalization by explicitly conditioning the policy on a parametric context space. In this paper, we further structure the contextual policy representation. We propose to factor contexts into two components: target contexts that describe the task objectives, e.g. target position for throwing a ball; and environment contexts that characterize the environment, e.g. initial position or mass of the ball. Our key observation is that experience can be directly generalized over target contexts. We show that this can be easily exploited in contextual policy search algorithms. In particular, we apply factorization to a Bayesian optimization approach to contextual policy search both in sampling-based and active learning settings. Our simulation results show faster learning and better generalization in various robotic domains. See our supplementary video: https://youtu.be/MNTbBAOufDY.

arXiv:1904.10642

Towards Combining On-Off-Policy Methods for Real-World Applications

In this paper, we point out a fundamental property of the objective in reinforcement learning, with which we can reformulate the policy gradient objective into a perceptron-like loss function, removing the need to distinguish between on and off policy training. Namely, we posit that it is sufficient to only update a policy $\pi$ for cases that satisfy the condition $A(\frac{\pi}{\mu}-1)\leq0$, where $A$ is the advantage, and $\mu$ is another policy. Furthermore, we show via theoretic derivation that a perceptron-like loss function matches the clipped surrogate objective for PPO. With our new formulation, the policies $\pi$ and $\mu$ can be arbitrarily apart in theory, effectively enabling off-policy training. To examine our derivations, we can combine the on-policy PPO clipped surrogate (which we show to be equivalent with one instance of the new reformation) with the off-policy IMPALA method. We first verify the combined method on the OpenAI Gym pendulum toy problem. Next, we use our method to train a quadrotor position controller in a simulator. Our trained policy is efficient and lightweight enough to perform in a low cost micro-controller at a minimum update rate of 500 Hz. For the quadrotor, we show two experiments to verify our method and demonstrate performance: 1) hovering at a fixed position, and 2) tracking along a specific trajectory. In preliminary trials, we are also able to apply the method to a real-world quadrotor.

arXiv:1904.10613

Defocused images removal of axial overlapping scattering particles by using three-dimensional nonlinear diffusion based on digital holography

We propose a three-dimensional nonlinear diffusion method to implement the similar autofocusing function of multiple micro-objects and simultaneously remove the defocused images, which can distinguish the locations of certain sized scattering particles that are overlapping along z-axis. It is applied to all of the reconstruction slices that are generated from the captured hologram after each back propagation. For certain small sized particles, the maxima of maximum gradient magnitude of each reconstruction slice appears at the ground truth z position after applying the proposed scheme when the reconstruction range along z-axis is sufficiently long and the reconstruction depth spacing is sufficiently fine. Therefore, the reconstructed image at ground truth z position is remained, while the defocused images are diffused out. The results demonstrated that the proposed scheme can diffuse out the defocused images which are 20 um away from the ground truth z position in spite of that several scattering particles with different diameters are completely overlapping along z-axis with a distance of 800 um when the hologram pixel pitch is 2 um. It also demonstrated that the sparsity distribution of the ground truth z slice cannot be affected by the sparsity distribution of corresponding defocused images when the diameter of the particle is not more than 35um and the reconstruction depth spacing is not less than 20 um.

arXiv:1904.10405

Big Math and the One-Brain Barrier A Position Paper and Architecture Proposal

Over the last decades, a class of important mathematical results have required an ever increasing amount of human effort to carry out. For some, the help of computers is now indispensable. We analyze the implications of this trend towards "big mathematics", its relation to human cognition, and how machine support for big math can be organized. The central contribution of this position paper is an information model for "doing mathematics", which posits that humans very efficiently integrate four aspects: inference, computation, tabulation, and narration around a well-organized core of mathematical knowledge. The challenge for mathematical software systems is that these four aspects need to be integrated as well. We briefly survey the state of the art.

arXiv:1904.10273공식 코드 · Papers with Code

Sequential modeling of Sessions using Recurrent Neural Networks for Skip Prediction

Recommender systems play an essential role in music streaming services, prominently in the form of personalized playlists. Exploring the user interactions within these listening sessions can be beneficial to understanding the user preferences in the context of a single session. In the 'Spotify Sequential Skip Prediction Challenge', WSDM, and Spotify are challenging people to understand the way users sequentially interact with music. We describe our solution approach in this paper and also state proposals for further improvements to the model. The proposed model initially generates a fixed vector representation of the session, and this additional information is incorporated into an Encoder-Decoder style architecture. This method achieved the seventh position in the competition, with a mean average accuracy of 0.604 on the test set. The solution code is available at https://github.com/sainathadapa/spotify-sequential-skip-prediction.

arXiv:1904.09599

A new insight into the Position Optimization of Wave Energy Converters by a Hybrid Local Search

Renewable energy, such as ocean wave energy, plays a pivotal role in addressing the tremendous growth of global energy demand. It is expected that wave energy will be one of the fastest-growing energy resources in the next decade, offering an enormous potential source of sustainable energy. This research investigates the placement optimization of oscillating buoy-type wave energy converters (WEC). The design of a wave farm consisting of an array of fully submerged three-tether buoys is evaluated. In a wave farm, buoy positions have a notable impact on the farm's output. Optimizing the buoy positions is a challenging research problem because of very complex interactions (constructive and destructive) between buoys. The main purpose of this research is maximizing the power output of the farm through the placement of buoys in a size-constrained environment. This paper proposes a new hybrid approach of the heuristic local search combined with a numerical optimization method that utilizes a knowledge-based surrogate power model. We compare the proposed hybrid method with other state-of-the-art search methods in five different wave scenarios -- one simplified irregular wave model and four real wave climates. Our method considerably outperforms all previous heuristic methods in terms of both quality of achieved solutions and the convergence-rate of search in all tested wave regimes.

arXiv:1904.08170

CaseNet: Content-Adaptive Scale Interaction Networks for Scene Parsing

Objects at different spatial positions in an image exhibit different scales. Adaptive receptive fields are expected to capture suitable ranges of context for accurate pixel level semantic prediction. Recently, atrous convolution with different dilation rates has been used to generate features of multi-scales through several branches which are then fused for prediction. However, there is a lack of explicit interaction among the branches of different scales to adaptively make full use of the contexts. In this paper, we propose a Content-Adaptive Scale Interaction Network (CASINet) to exploit the multi-scale features for scene parsing. We build CASINet based on the classic Atrous Spatial Pyramid Pooling (ASPP) module, followed by a proposed contextual scale interaction (CSI) module, and a scale adaptation (SA) module. Specifically, in the CSI module, for each spatial position of some scale, instead of being limited by a fixed set of convolutional filters that are shared across different spatial positions for feature learning, we promote the adaptivity of the convolutional filters to spatial positions. We achieve this by the context interaction among the features of different scales. The SA module explicitly and softly selects the suitable scale for each spatial position and each channel. Ablation studies demonstrate the effectiveness of the proposed modules. We achieve state-of-the-art performance on three scene parsing benchmarks Cityscapes, ADE20K and LIP.

arXiv:1904.08010

How to define co-occurrence in different domains of study?

This position paper presents a comparative study of co-occurrences. Some similarities and differences in the definition exist depending on the research domain (e.g. linguistics, NLP, computer science). This paper discusses these points, and deals with the methodological aspects in order to identify co-occurrences in a multidisciplinary paradigm.

arXiv:1904.07361

Custom Video-Oculography Device and Its Application to Fourth Purkinje Image Detection during Saccades

We built a custom video-based eye-tracker that saves every video frame as a full resolution image (MJPEG). Images can be processed offline for the detection of ocular features, including the pupil and corneal reflection (First Purkinje Image, P1) position. A comparison of multiple algorithms for detection of pupil and corneal reflection can be performed. The system provides for highly flexible stimulus creation, with mixing of graphic, image, and video stimuli. We can change cameras and infrared illuminators depending on the image qualities and frame rate desired. Using this system, we have detected the position of the Fourth Purkinje image (P4) in the frames. We show that when we estimate gaze by calculating P1-P4, signal compares well with gaze estimated with a DPI eye-tracker, which natively detects and tracks the P1 and P4.

arXiv:1904.07776

The ALOS Dataset for Advert Localization in Outdoor Scenes

The rapid increase in the number of online videos provides the marketing and advertising agents ample opportunities to reach out to their audience. One of the most widely used strategies is product placement, or embedded marketing, wherein new advertisements are integrated seamlessly into existing advertisements in videos. Such strategies involve accurately localizing the position of the advert in the image frame, either manually in the video editing phase, or by using machine learning frameworks. However, these machine learning techniques and deep neural networks need a massive amount of data for training. In this paper, we propose and release the first large-scale dataset of advertisement billboards, captured in outdoor scenes. We also benchmark several state-of-the-art semantic segmentation algorithms on our proposed dataset.

arXiv:1904.06501

Maximum Correntropy Criterion with Variable Center

Correntropy is a local similarity measure defined in kernel space and the maximum correntropy criterion (MCC) has been successfully applied in many areas of signal processing and machine learning in recent years. The kernel function in correntropy is usually restricted to the Gaussian function with center located at zero. However, zero-mean Gaussian function may not be a good choice for many practical applications. In this study, we propose an extended version of correntropy, whose center can locate at any position. Accordingly, we propose a new optimization criterion called maximum correntropy criterion with variable center (MCC-VC). We also propose an efficient approach to optimize the kernel width and center location in MCC-VC. Simulation results of regression with linear in parameters (LIP) models confirm the desirable performance of the new method.

arXiv:1904.07099

Processsing Simple Geometric Attributes with Autoencoders

Image synthesis is a core problem in modern deep learning, and many recent architectures such as autoencoders and Generative Adversarial networks produce spectacular results on highly complex data, such as images of faces or landscapes. While these results open up a wide range of new, advanced synthesis applications, there is also a severe lack of theoretical understanding of how these networks work. This results in a wide range of practical problems, such as difficulties in training, the tendency to sample images with little or no variability, and generalisation problems. In this paper, we propose to analyse the ability of the simplest generative network, the autoencoder, to encode and decode two simple geometric attributes : size and position. We believe that, in order to understand more complicated tasks, it is necessary to first understand how these networks process simple attributes. For the first property, we analyse the case of images of centred disks with variable radii. We explain how the autoencoder projects these images to and from a latent space of smallest possible dimension, a scalar. In particular, we describe a closed-form solution to the decoding training problem in a network without biases, and show that during training, the network indeed finds this solution. We then investigate the best regularisation approaches which yield networks that generalise well. For the second property, position, we look at the encoding and decoding of Dirac delta functions, also known as `one-hot' vectors. We describe a hand-crafted filter that achieves encoding perfectly, and show that the network naturally finds this filter during training. We also show experimentally that the decoding can be achieved if the dataset is sampled in an appropriate manner.

arXiv:1904.06187

Position-Aware Convolutional Networks for Traffic Prediction

Forecasting the future traffic flow distribution in an area is an important issue for traffic management in an intelligent transportation system. The key challenge of traffic prediction is to capture spatial and temporal relations between future traffic flows and historical traffic due to highly dynamical patterns of human activities. Most existing methods explore such relations by fusing spatial and temporal features extracted from multi-source data. However, they neglect position information which helps distinguish patterns on different positions. In this paper, we propose a position-aware neural network that integrates data features and position information. Our approach employs the inception backbone network to capture rich features of traffic distribution on the whole area. The novelty lies in that under the backbone network, we apply position embedding technique used in neural language processing to represent position information as embedding vectors which are learned during the training. With these embedding vectors, we design position-aware convolution which allows different kernels to process features of different positions. Extensive experiments on two real-world datasets show that our approach outperforms previous methods even with fewer data sources.

arXiv:1904.04912

Enhancing Time Series Momentum Strategies Using Deep Neural Networks

While time series momentum is a well-studied phenomenon in finance, common strategies require the explicit definition of both a trend estimator and a position sizing rule. In this paper, we introduce Deep Momentum Networks -- a hybrid approach which injects deep learning based trading rules into the volatility scaling framework of time series momentum. The model also simultaneously learns both trend estimation and position sizing in a data-driven manner, with networks directly trained by optimising the Sharpe ratio of the signal. Backtesting on a portfolio of 88 continuous futures contracts, we demonstrate that the Sharpe-optimised LSTM improved traditional methods by more than two times in the absence of transactions costs, and continue outperforming when considering transaction costs up to 2-3 basis points. To account for more illiquid assets, we also propose a turnover regularisation term which trains the network to factor in costs at run-time.

arXiv:1904.04866공식 코드 · Papers with Code

Characterizing the impact of geometric properties of word embeddings on task performance

Analysis of word embedding properties to inform their use in downstream NLP tasks has largely been studied by assessing nearest neighbors. However, geometric properties of the continuous feature space contribute directly to the use of embedding features in downstream models, and are largely unexplored. We consider four properties of word embedding geometry, namely: position relative to the origin, distribution of features in the vector space, global pairwise distances, and local pairwise distances. We define a sequence of transformations to generate new embeddings that expose subsets of these properties to downstream models and evaluate change in task performance to understand the contribution of each property to NLP models. We transform publicly available pretrained embeddings from three popular toolkits (word2vec, GloVe, and FastText) and evaluate on a variety of intrinsic tasks, which model linguistic information in the vector space, and extrinsic tasks, which use vectors as input to machine learning models. We find that intrinsic evaluations are highly sensitive to absolute position, while extrinsic tasks rely primarily on local similarity. Our findings suggest that future embedding models and post-processing techniques should focus primarily on similarity to nearby points in vector space.