PyTorchKR Papers

최신 논문

총 3,382편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:1804.06909

Modeling and Simultaneously Removing Bias via Adversarial Neural Networks

In real world systems, the predictions of deployed Machine Learned models affect the training data available to build subsequent models. This introduces a bias in the training data that needs to be addressed. Existing solutions to this problem attempt to resolve the problem by either casting this in the reinforcement learning framework or by quantifying the bias and re-weighting the loss functions. In this work, we develop a novel Adversarial Neural Network (ANN) model, an alternative approach which creates a representation of the data that is invariant to the bias. We take the Paid Search auction as our working example and ad display position features as the confounding features for this setting. We show the success of this approach empirically on both synthetic data as well as real world paid search auction data from a major search engine.

arXiv:1804.04603

Outline Objects using Deep Reinforcement Learning

Image segmentation needs both local boundary position information and global object context information. The performance of the recent state-of-the-art method, fully convolutional networks, reaches a bottleneck due to the neural network limit after balancing between the two types of information simultaneously in an end-to-end training style. To overcome this problem, we divide the semantic image segmentation into temporal subtasks. First, we find a possible pixel position of some object boundary; then trace the boundary at steps within a limited length until the whole object is outlined. We present the first deep reinforcement learning approach to semantic image segmentation, called DeepOutline, which outperforms other algorithms in Coco detection leaderboard in the middle and large size person category in Coco val2017 dataset. Meanwhile, it provides an insight into a divide and conquer way by reinforcement learning on computer vision problems.

arXiv:1804.08266공식 코드 · Papers with Code

NLITrans at SemEval-2018 Task 12: Transfer of Semantic Knowledge for Argument Comprehension

The Argument Reasoning Comprehension Task requires significant language understanding and complex reasoning over world knowledge. We focus on transfer of a sentence encoder to bootstrap more complicated models given the small size of the dataset. Our best model uses a pre-trained BiLSTM to encode input sentences, learns task-specific features for the argument and warrants, then performs independent argument-warrant matching. This model achieves mean test set accuracy of 64.43%. Encoder transfer yields a significant gain to our best model over random initialization. Independent warrant matching effectively doubles the size of the dataset and provides additional regularization. We demonstrate that regularization comes from ignoring statistical correlations between warrant features and position. We also report an experiment with our best model that only matches warrants to reasons, ignoring claims. Relatively low performance degradation suggests that our model is not necessarily learning the intended task.

arXiv:1804.08243

Constructing Locally Dense Point Clouds Using OpenSfM and ORB-SLAM2

This paper aims at finding a method to register two different point clouds constructed by ORB-SLAM2 and OpenSfM. To do this, we post some tags with unique textures in the scene and take videos and photos of that area. Then we take short videos of only the tags to extract their features. By matching the ORB feature of the tags with their corresponding features in the scene, it is then possible to localize the position of these tags both in point clouds constructed by ORB-SLAM2 and OpenSfM. Thus, the best transformation matrix between two point clouds can be calculated, and the two point clouds can be aligned.

arXiv:1707.02920공식 코드 · Papers with Code

Vision-Based Multi-Task Manipulation for Inexpensive Robots Using End-To-End Learning from Demonstration

We propose a technique for multi-task learning from demonstration that trains the controller of a low-cost robotic arm to accomplish several complex picking and placing tasks, as well as non-prehensile manipulation. The controller is a recurrent neural network using raw images as input and generating robot arm trajectories, with the parameters shared across the tasks. The controller also combines VAE-GAN-based reconstruction with autoregressive multimodal action prediction. Our results demonstrate that it is possible to learn complex manipulation tasks, such as picking up a towel, wiping an object, and depositing the towel to its previous position, entirely from raw images with direct behavior cloning. We show that weight sharing and reconstruction-based regularization substantially improve generalization and robustness, and training on multiple tasks simultaneously increases the success rate on all tasks.

arXiv:1804.09154

DOOM Level Generation using Generative Adversarial Networks

We applied Generative Adversarial Networks (GANs) to learn a model of DOOM levels from human-designed content. Initially, we analysed the levels and extracted several topological features. Then, for each level, we extracted a set of images identifying the occupied area, the height map, the walls, and the position of game objects. We trained two GANs: one using plain level images, one using both the images and some of the features extracted during the preliminary analysis. We used the two networks to generate new levels and compared the results to assess whether the network trained using also the topological features could generate levels more similar to human-designed ones. Our results show that GANs can capture intrinsic structure of DOOM levels and appears to be a promising approach to level generation in first person shooter games.

arXiv:1804.10500

Deep Reinforcement Learning to Acquire Navigation Skills for Wheel-Legged Robots in Complex Environments

Mobile robot navigation in complex and dynamic environments is a challenging but important problem. Reinforcement learning approaches fail to solve these tasks efficiently due to reward sparsities, temporal complexities and high-dimensionality of sensorimotor spaces which are inherent in such problems. We present a novel approach to train action policies to acquire navigation skills for wheel-legged robots using deep reinforcement learning. The policy maps height-map image observations to motor commands to navigate to a target position while avoiding obstacles. We propose to acquire the multifaceted navigation skill by learning and exploiting a number of manageable navigation behaviors. We also introduce a domain randomization technique to improve the versatility of the training samples. We demonstrate experimentally a significant improvement in terms of data-efficiency, success rate, robustness against irrelevant sensory data, and also the quality of the maneuver skills.

arXiv:1804.10361

An Element Sensitive Saliency Model with Position Prior Learning for Web Pages

Understanding human visual attention is important for multimedia applications. Many studies have attempted to learn from eye-tracking data and build computational saliency prediction models. However, limited efforts have been devoted to saliency prediction for Web pages, which are characterized by more diverse content elements and spatial layouts. In this paper, we propose a novel end-to-end deep generative saliency model for Web pages. To capture position biases introduced by page layouts, a Position Prior Learning sub-network is proposed, which models position biases as multivariate Gaussian distribution using variational auto-encoder. To model different elements of a Web page, a Multi Discriminative Region Detection (MDRD) branch and a Text Region Detection(TRD) branch are introduced, which target to extract discriminative localizations and "prominent" text regions likely to correspond to human attention, respectively. We validate the proposed model with FiWI, a public Web-page dataset, and shows that the proposed model outperforms the state-of-art models for Web-page saliency prediction.

arXiv:1712.01208

The Case for Learned Index Structures

Indexes are models: a B-Tree-Index can be seen as a model to map a key to the position of a record within a sorted array, a Hash-Index as a model to map a key to a position of a record within an unsorted array, and a BitMap-Index as a model to indicate if a data record exists or not. In this exploratory research paper, we start from this premise and posit that all existing index structures can be replaced with other types of models, including deep-learning models, which we term learned indexes. The key idea is that a model can learn the sort order or structure of lookup keys and use this signal to effectively predict the position or existence of records. We theoretically analyze under which conditions learned indexes outperform traditional index structures and describe the main challenges in designing learned index structures. Our initial results show, that by using neural nets we are able to outperform cache-optimized B-Trees by up to 70% in speed while saving an order-of-magnitude in memory over several real-world data sets. More importantly though, we believe that the idea of replacing core components of a data management system through learned models has far reaching implications for future systems designs and that this work just provides a glimpse of what might be possible.

arXiv:1805.02241

Acquisition and use of knowledge over a restricted domain by intelligent agents

This short paper provides a description of an architecture to acquisition and use of knowledge by intelligent agents over a restricted domain of the Internet Infrastructure. The proposed architecture is added to an intelligent agent deployment model over a very useful server for Internet Autonomous System administrators. Such servers, which are heavily dependent on arbitrary and eventual updates of human beings, become unreliable. This is a position paper that proposes three research questions that are still in progress.

arXiv:1805.02020

Position Estimation of Camera Based on Unsupervised Learning

It is an exciting task to recover the scene's 3d-structure and camera pose from the video sequence. Most of the current solutions divide it into two parts, monocular depth recovery and camera pose estimation. The monocular depth recovery is often studied as an independent part, and a better depth estimation is used to solve the pose. While camera pose is still estimated by traditional SLAM (Simultaneous Localization And Mapping) methods in most cases. The use of unsupervised method for monocular depth recovery and pose estimation has benefited from the study of [1] and achieved good results. In this paper, we improve the method of [1]. Our emphasis is laid on the improvement of the idea and related theory, introducing a more reasonable inter frame constraints and finally synthesize the camera trajectory with inter frame pose estimation in the unified world coordinate system. And our results get better performance.

arXiv:1805.02823

Hierarchical Structured Model for Fine-to-coarse Manifesto Text Analysis

Election manifestos document the intentions, motives, and views of political parties. They are often used for analysing a party's fine-grained position on a particular issue, as well as for coarse-grained positioning of a party on the left--right spectrum. In this paper we propose a two-stage model for automatically performing both levels of analysis over manifestos. In the first step we employ a hierarchical multi-task structured deep model to predict fine- and coarse-grained positions, and in the second step we perform post-hoc calibration of coarse-grained positions using probabilistic soft logic. We empirically show that the proposed model outperforms state-of-art approaches at both granularities using manifestos from twelve countries, written in ten different languages.

arXiv:1805.03269

Sparse Blind Deconvolution for Distributed Radar Autofocus Imaging

A common problem that arises in radar imaging systems, especially those mounted on mobile platforms, is antenna position ambiguity. Approaches to resolve this ambiguity and correct position errors are generally known as radar autofocus. Common techniques that attempt to resolve the antenna ambiguity generally assume an unknown gain and phase error afflicting the radar measurements. However, ensuring identifiability and tractability of the unknown error imposes strict restrictions on the allowable antenna perturbations. Furthermore, these techniques are often not applicable in near-field imaging, where mapping the position ambiguity to phase errors breaks down. In this paper, we propose an alternate formulation where the position error of each antenna is mapped to a spatial shift operator in the image-domain. Thus, the radar autofocus problem becomes a multichannel blind deconvolution problem, in which the radar measurements correspond to observations of a static radar image that is convolved with the spatial shift kernel associated with each antenna. To solve the reformulated problem, we also develop a block coordinate descent framework that leverages the sparsity and piece-wise smoothness of the radar scene, as well as the one-sparse property of the two dimensional shift kernels. We evaluate the performance of our approach using both simulated and experimental radar measurements, and demonstrate its superior performance compared to state-of-the-art methods.

arXiv:1803.07244

The Three Pillars of Machine Programming

In this position paper, we describe our vision of the future of machine programming through a categorical examination of three pillars of research. Those pillars are: (i) intention, (ii) invention, and(iii) adaptation. Intention emphasizes advancements in the human-to-computer and computer-to-machine-learning interfaces. Invention emphasizes the creation or refinement of algorithms or core hardware and software building blocks through machine learning (ML). Adaptation emphasizes advances in the use of ML-based constructs to autonomously evolve software.

arXiv:1804.08813공식 코드 · Papers with Code

End-Task Oriented Textual Entailment via Deep Explorations of Inter-Sentence Interactions

This work deals with SciTail, a natural entailment challenge derived from a multi-choice question answering problem. The premises and hypotheses in SciTail were generated with no awareness of each other, and did not specifically aim at the entailment task. This makes it more challenging than other entailment data sets and more directly useful to the end-task -- question answering. We propose DEISTE (deep explorations of inter-sentence interactions for textual entailment) for this entailment task. Given word-to-word interactions between the premise-hypothesis pair ($P$, $H$), DEISTE consists of: (i) a parameter-dynamic convolution to make important words in $P$ and $H$ play a dominant role in learnt representations; and (ii) a position-aware attentive convolution to encode the representation and position information of the aligned word pairs. Experiments show that DEISTE gets $\approx$5\% improvement over prior state of the art and that the pretrained DEISTE on SciTail generalizes well on RTE-5.

arXiv:1805.06485공식 코드 · Papers with Code

QuaterNet: A Quaternion-based Recurrent Model for Human Motion

Deep learning for predicting or generating 3D human pose sequences is an active research area. Previous work regresses either joint rotations or joint positions. The former strategy is prone to error accumulation along the kinematic chain, as well as discontinuities when using Euler angle or exponential map parameterizations. The latter requires re-projection onto skeleton constraints to avoid bone stretching and invalid configurations. This work addresses both limitations. Our recurrent network, QuaterNet, represents rotations with quaternions and our loss function performs forward kinematics on a skeleton to penalize absolute position errors instead of angle errors. On short-term predictions, QuaterNet improves the state-of-the-art quantitatively. For long-term generation, our approach is qualitatively judged as realistic as recent neural strategies from the graphics literature.

arXiv:1805.06406

Deep Segmentation and Registration in X-Ray Angiography Video

In interventional radiology, short video sequences of vein structure in motion are captured in order to help medical personnel identify vascular issues or plan intervention. Semantic segmentation can greatly improve the usefulness of these videos by indicating exact position of vessels and instruments, thus reducing the ambiguity. We propose a real-time segmentation method for these tasks, based on U-Net network trained in a Siamese architecture from automatically generated annotations. We make use of noisy low level binary segmentation and optical flow to generate multi class annotations that are successively improved in a multistage segmentation approach. We significantly improve the performance of a state of the art U-Net at the processing speeds of 90fps.

arXiv:1805.10799공식 코드 · Papers with Code

Interactive Text2Pickup Network for Natural Language based Human-Robot Collaboration

In this paper, we propose the Interactive Text2Pickup (IT2P) network for human-robot collaboration which enables an effective interaction with a human user despite the ambiguity in user's commands. We focus on the task where a robot is expected to pick up an object instructed by a human, and to interact with the human when the given instruction is vague. The proposed network understands the command from the human user and estimates the position of the desired object first. To handle the inherent ambiguity in human language commands, a suitable question which can resolve the ambiguity is generated. The user's answer to the question is combined with the initial command and given back to the network, resulting in more accurate estimation. The experiment results show that given unambiguous commands, the proposed method can estimate the position of the requested object with an accuracy of 98.49% based on our test dataset. Given ambiguous language commands, we show that the accuracy of the pick up task increases by 1.94 times after incorporating the information obtained from the interaction.

arXiv:1710.06473

Superpixels Based Marker Tracking Vs. Hue Thresholding In Rodent Biomechanics Application

Examining locomotion has improved our basic understanding of motor control and aided in treating motor impairment. Mice and rats are premier models of human disease and increasingly the model systems of choice for basic neuroscience. High frame rates (250 Hz) are needed to quantify the kinematics of these running rodents. Manual tracking, especially for multiple markers, becomes time-consuming and impossible for large sample sizes. Therefore, the need for automatic segmentation of these markers has grown in recent years. We propose two methods to segment and track these markers: first, using SLIC superpixels segmentation with a tracker based on position, speed, shape, and color information of the segmented region in the previous frame; second, using a thresholding on hue channel following up with the same tracker. The comparison showed that the SLIC superpixels method was superior because the segmentation was more reliable and based on both color and spatial information.

arXiv:1801.02025

Robot Localisation and 3D Position Estimation Using a Free-Moving Camera and Cascaded Convolutional Neural Networks

Many works in collaborative robotics and human-robot interaction focuses on identifying and predicting human behaviour while considering the information about the robot itself as given. This can be the case when sensors and the robot are calibrated in relation to each other and often the reconfiguration of the system is not possible, or extra manual work is required. We present a deep learning based approach to remove the constraint of having the need for the robot and the vision sensor to be fixed and calibrated in relation to each other. The system learns the visual cues of the robot body and is able to localise it, as well as estimate the position of robot joints in 3D space by just using a 2D color image. The method uses a cascaded convolutional neural network, and we present the structure of the network, describe our own collected dataset, explain the network training and achieved results. A fully trained system shows promising results in providing an accurate mask of where the robot is located and a good estimate of its joints positions in 3D. The accuracy is not good enough for visual servoing applications yet, however, it can be sufficient for general safety and some collaborative tasks not requiring very high precision. The main benefit of our method is the possibility of the vision sensor to move freely. This allows it to be mounted on moving objects, for example, a body of the person or a mobile robot working in the same environment as the robots are operating in.