PyTorchKR Papers

최신 논문

총 3,337편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2107.03056

Position Constrained, Adaptive Control of Robotic Manipulators without Velocity Measurements

This work presents the design and the corresponding stability analysis of a model based, joint position tracking error constrained, adaptive output feedback controller for robot manipulators. Specifically, provided that the initial joint position tracking error starts within a predefined region, the proposed controller algorithm ensures that the joint tracking error remains inside this region and asymptotically approaches to zero, despite the lack of joint velocity measurements and uncertainties associated with the system dynamics. The need for the joint velocity measurements are removed via the use of a surrogate filter formulation in conjunction with the use of desired model compensation. The stability and the convergence of the closed loop system are proved via a barrier Lyapunov function based argument. A simulation performed on a two-link robotic manipulator is provided in order to illustrate the feasibility and effectiveness of the proposed method.

arXiv:2107.02984

Deep Convolutional Correlation Iterative Particle Filter for Visual Tracking

This work proposes a novel framework for visual tracking based on the integration of an iterative particle filter, a deep convolutional neural network, and a correlation filter. The iterative particle filter enables the particles to correct themselves and converge to the correct target position. We employ a novel strategy to assess the likelihood of the particles after the iterations by applying K-means clustering. Our approach ensures a consistent support for the posterior distribution. Thus, we do not need to perform resampling at every video frame, improving the utilization of prior distribution information. Experimental results on two different benchmark datasets show that our tracker performs favorably against state-of-the-art methods.

arXiv:2107.02444

The NiuTrans End-to-End Speech Translation System for IWSLT 2021 Offline Task

This paper describes the submission of the NiuTrans end-to-end speech translation system for the IWSLT 2021 offline task, which translates from the English audio to German text directly without intermediate transcription. We use the Transformer-based model architecture and enhance it by Conformer, relative position encoding, and stacked acoustic and textual encoding. To augment the training data, the English transcriptions are translated to German translations. Finally, we employ ensemble decoding to integrate the predictions from several models trained with the different datasets. Combining these techniques, we achieve 33.84 BLEU points on the MuST-C En-De test set, which shows the enormous potential of the end-to-end model.

arXiv:2107.02239공식 코드 · Papers with Code

Vision Xformers: Efficient Attention for Image Classification

Although transformers have become the neural architectures of choice for natural language processing, they require orders of magnitude more training data, GPU memory, and computations in order to compete with convolutional neural networks for computer vision. The attention mechanism of transformers scales quadratically with the length of the input sequence, and unrolled images have long sequence lengths. Plus, transformers lack an inductive bias that is appropriate for images. We tested three modifications to vision transformer (ViT) architectures that address these shortcomings. Firstly, we alleviate the quadratic bottleneck by using linear attention mechanisms, called X-formers (such that, X in {Performer, Linformer, Nystr\"omformer}), thereby creating Vision X-formers (ViXs). This resulted in up to a seven times reduction in the GPU memory requirement. We also compared their performance with FNet and multi-layer perceptron mixers, which further reduced the GPU memory requirement. Secondly, we introduced an inductive bias for images by replacing the initial linear embedding layer by convolutional layers in ViX, which significantly increased classification accuracy without increasing the model size. Thirdly, we replaced the learnable 1D position embeddings in ViT with Rotary Position Embedding (RoPE), which increases the classification accuracy for the same model size. We believe that incorporating such changes can democratize transformers by making them accessible to those with limited data and computing resources.

arXiv:2107.02155

Reconfigurable Intelligent Surface Assisted Device-to-Device Communications

Reconfigurable intelligent surface (RIS) technology is a promising method to enhance wireless communications services and to realize the smart radio environment. In this paper, we investigate the application of RIS in D2D communications, and maximize the sum of the transmission rate of the D2D underlaying networks in a new perspective. Instead of solving similarly formulated resource allocation problems for D2D communications, this paper treats the wireless environment as a variable by adjusting the position and phase shift of the RIS. To solve this non-convex problem, we propose a novel double deep Q-network (DDQN) based structure which is able to achieve the near-optimal performance with lower complexity and enhanced robustness. Simulation results illustrate that the proposed DDQN based structure can achieve a higher uplink rate compared to the benchmarks, meanwhile meeting the quality of service (QoS) requirements at the base station (BS) and D2D receivers.

arXiv:2107.01379

Vestibular Drop Attacks and Meniere's Disease as Results of Otolithic Membrane Damage -- A Numerical Model

BACKGROUND: Meniere's Disease (MD) is a condition of the inner ear with symptoms affecting both vestibular and hearing functions. Some patients with MD experience vestibular drop attacks (VDAs), which are violent falls caused by spurious vestibular signals from the utricle and/or saccule. Recent surgical work has shown that patients who experience VDAs also show distrupted utricular otolithic membranes. OBJECTIVE: The objective of this study is to determine if otolithic membrane damage alone is sufficient to induce spurious vestibular signals, thus potentially eliciting VDAs and the vestibular dysfunction seen in patients with MD. METHODS: We use a previously developed numerical model to describe the nonlinear dynamics of an array of active, elastically coupled hair cells. We then reduce the coupling strength of a selected region of the membrane to model the effects of tissue damage. RESULTS: As we reduce the coupling strength, we observe large and abrupt spikes in hair bundle position. As bundle displacements from the equilibrium position have been shown to lead to depolarization of the hair-cell soma and hence trigger neural activity, this spontaneous activity could elicit false detection of a vestibular signal. CONCLUSIONS: The results of this numerical model suggest that otolithic membrane damage alone may be sufficient to induce VDAs and the vestibular dysfunction seen in patients with MD. Future experimental work is needed to confirm these results in vitro.

arXiv:2107.01002

WiCluster: Passive Indoor 2D/3D Positioning using WiFi without Precise Labels

We introduce WiCluster, a new machine learning (ML) approach for passive indoor positioning using radio frequency (RF) channel state information (CSI). WiCluster can predict both a zone-level position and a precise 2D or 3D position, without using any precise position labels during training. Prior CSI-based indoor positioning work has relied on non-parametric approaches using digital signal-processing (DSP) and, more recently, parametric approaches (e.g., fully supervised ML methods). However these do not handle the complexity of real-world environments well and do not meet requirements for large-scale commercial deployments: the accuracy of DSP-based method deteriorates significantly in non-line-of-sight conditions, while supervised ML methods need large amounts of hard-to-acquire centimeter accuracy position labels. In contrast, WiCluster is precise, requires weaker label-information that can be easily collected, and works well in non-line-of-sight conditions. Our first contribution is a novel dimensionality reduction method for charting. It combines a triplet-loss with a multi-scale clustering-loss to map the high-dimensional CSI representation to a 2D/3D latent space. Our second contribution is two weakly supervised losses that map this latent space into a Cartesian map, resulting in meter-accuracy position results. These losses only require simple to acquire priors: a sketch of the floorplan, approximate access-point locations and a few CSI packets that are labelled with the corresponding zone in the floorplan. Thirdly, we report results and a robustness study for 2D positioning in two single-floor office buildings and 3D positioning in a two-story home.

arXiv:2107.00761공식 코드 · Papers with Code

On the Bike Spreading Problem

A free-floating bike-sharing system (FFBSS) is a dockless rental system where an individual can borrow a bike and returns it anywhere, within the service area. To improve the rental service, available bikes should be distributed over the entire service area: a customer leaving from any position is then more likely to find a near bike and then to use the service. Moreover, spreading bikes among the entire service area increases urban spatial equity since the benefits of FFBSS are not a prerogative of just a few zones. For guaranteeing such distribution, the FFBSS operator can use vans to manually relocate bikes, but it incurs high economic and environmental costs. We propose a novel approach that exploits the existing bike flows generated by customers to distribute bikes. More specifically, by envisioning the problem as an Influence Maximization problem, we show that it is possible to position batches of bikes on a small number of zones, and then the daily use of FFBSS will efficiently spread these bikes on a large area. We show that detecting these zones is NP-complete, but there exists a simple and efficient $1-1/e$ approximation algorithm; our approach is then evaluated on a dataset of rides from the free-floating bike-sharing system of the city of Padova.

arXiv:2107.00692

Interactive decoding of words from visual speech recognition models

This work describes an interactive decoding method to improve the performance of visual speech recognition systems using user input to compensate for the inherent ambiguity of the task. Unlike most phoneme-to-word decoding pipelines, which produce phonemes and feed these through a finite state transducer, our method instead expands words in lockstep, facilitating the insertion of interaction points at each word position. Interaction points enable us to solicit input during decoding, allowing users to interactively direct the decoding process. We simulate the behavior of user input using an oracle to give an automated evaluation, and show promise for the use of this method for text input.

arXiv:2107.00962공식 코드 · Papers with Code

Brain over Brawn: Using a Stereo Camera to Detect, Track, and Intercept a Faster UAV by Reconstructing the Intruder's Trajectory

This paper presents our approach to intercepting a faster intruder UAV, inspired by the MBZIRC 2020 Challenge 1. By utilizing a priori knowledge of the shape of the intruder's trajectory, we can calculate an interception point. Target tracking is based on image processing by a YOLOv3 Tiny convolutional neural network, combined with depth calculation using a gimbal-mounted ZED Mini stereo camera. We use RGB and depth data from the camera, devising a noise-reducing histogram-filter to extract the target's 3D position. Obtained 3D measurements of target's position are used to calculate the position, orientation, and size of a figure-eight shaped trajectory, which we approximate using a Bernoulli lemniscate. Once the approximation is deemed sufficiently precise, as measured by the distance between observations and estimate, we calculate an interception point to position the interceptor UAV directly on the intruder's path. Our method, which we have significantly improved based on the experience gathered during the MBZIRC competition, has been validated in simulation and through field experiments. Our results confirm that we have developed an efficient, visual-perception module that can extract information describing the intruder UAV's motion with precision sufficient to support interception planning. In a majority of our simulated encounters, we can track and intercept a target that moves 30% faster than the interceptor. Corresponding tests in an unstructured environment yielded 9 out of 12 successful results.

arXiv:2107.00689

Aerial Map-Based Navigation Using Semantic Segmentation and Pattern Matching

This paper proposes a novel approach to map-based navigation system for unmanned aircraft. The proposed system attempts label-to-label matching, not image-to-image matching, between aerial images and a map database. The ground objects can be labelled by deep learning approaches and the configuration of the objects is used to find the corresponding location in the map database. The use of the deep learning technique as a tool for extracting high-level features reduces the image-based localization problem to a pattern matching problem. This paper proposes a pattern matching algorithm that does not require altitude information or a camera model to estimate the absolute horizontal position. The feasibility analysis with simulated images shows the proposed map-based navigation can be realized with the proposed pattern matching algorithm and it is able to provide positions given the labelled objects.

arXiv:2107.00337

PoliTO-IIT Submission to the EPIC-KITCHENS-100 Unsupervised Domain Adaptation Challenge for Action Recognition

In this report, we describe the technical details of our submission to the EPIC-Kitchens-100 Unsupervised Domain Adaptation (UDA) Challenge in Action Recognition. To tackle the domain-shift which exists under the UDA setting, we first exploited a recent Domain Generalization (DG) technique, called Relative Norm Alignment (RNA). It consists in designing a model able to generalize well to any unseen domain, regardless of the possibility to access target data at training time. Then, in a second phase, we extended the approach to work on unlabelled target data, allowing the model to adapt to the target distribution in an unsupervised fashion. For this purpose, we included in our framework existing UDA algorithms, such as Temporal Attentive Adversarial Adaptation Network (TA3N), jointly with new multi-stream consistency losses, namely Temporal Hard Norm Alignment (T-HNA) and Min-Entropy Consistency (MEC). Our submission (entry 'plnet') is visible on the leaderboard and it achieved the 1st position for 'verb', and the 3rd position for both 'noun' and 'action'.

arXiv:2106.15244

Bubble lifetimes in DNA gene promoters and their mutations affecting transcription

Relative lifetimes of inherent double stranded DNA openings with lengths up to ten base pairs are presented for different gene promoters and corresponding mutants that either increase or decrease transcriptional activity, in the framework of the Peyrard-Bishop-Dauxois model. Extensive microcanonical simulations are used, with energies corresponding to physiological temperature. The bubble lifetime profiles along the DNA sequences demonstrate a significant reduction of the average lifetime at the mutation sites when the mutated promoter decreases transcription, while a corresponding enhancement of the bubble lifetime is observed in the case of mutations leading to increased transcription. The relative difference of bubble lifetimes between the mutated and the wild type promoters at the position of mutation varies from 20% to more than 30% as the bubble length is decreasing

arXiv:2106.15134공식 코드 · Papers with Code

Analysis and Control of a Planar Quadrotor

In this paper, we model the planar motion of a quadcopter, and develop a linear model of the same. We perform stability analysis of the open loop system and develop a PD controller for its position control. We compare the closed loop response between the linear and non linear systems using the controller developed for the linear system. We also perform stability analysis for the linear and non linear systems, and compare the PD controller with modern deep neural methods based on reinforcement learning actor-critic networks.

arXiv:2106.15539

Model-Centric Volumetric Point Cloud Attributes

Point clouds have recently gained interest, especially for real-time applications and for 3D-scanned material, such as is used in autonomous driving, architecture, and engineering, to model real estate for renovation or display. Point clouds are associated with geometry information and attributes such as color. Be the color unique or direction-dependent (in the case of plenoptic point clouds), it reflects the colors observed by cameras displaced around the object. Hence, not only are the viewing references assumed, but the illumination spectrum and illumination geometry is also implicit. We propose a model-centric description of the 3D object, that is independent of the illumination and of the position of the cameras. We want to be able to describe the objects themselves such that, at a later stage, the rendering of the model may decide where to place illumination, from which it may calculate the image viewed by a given camera. We want to be able to describe transparent or translucid objects, mirrors, fishbowls, fog and smoke. Volumetric clouds may allow us to describe the air, however ``empty'', and introduce air particles, in a manner independent of the viewer position. For that, we rely on some eletromagnetic properties to arrive at seven attributes per voxel that would describe the material and its color or transparency. Three attributes are for the transmissivity of each color, three are for the attenuation of each color, and another attribute is for diffuseness. These attributes give information about the object to the renderer, with whom lies the decision on how to render and depict each object.

arXiv:2106.15117공식 코드 · Papers with Code

SDL: New data generation tools for full-level annotated document layout

We present a novel data generation tool for document processing. The tool focuses on providing a maximal level of visual information in a normal type document, ranging from character position to paragraph-level position. It also enables working with a large dataset on low-resource languages as well as providing a mean of processing thorough full-level information of the documented text. The data generation tools come with a dataset of 320000 Vietnamese synthetic document images and an instruction to generate a dataset of similar size in other languages. The repository can be found at: https://github.com/tson1997/SDL-Document-Image-Generation

arXiv:2106.15178

Deep Inertial Navigation using Continuous Domain Adaptation and Optimal Transport

In this paper, we propose a new strategy for learning inertial robotic navigation models. The proposed strategy enhances the generalisability of end-to-end inertial modelling, and is aimed at wheeled robotic deployments. Concretely, the paper describes the following. (1) Using precision robotics, we empirically characterise the effect of changing the sensor position during navigation on the distribution of raw inertial signals, as well as the corresponding impact on learnt latent spaces. (2) We propose neural architectures and algorithms to assimilate knowledge from an indexed set of sensor positions in order to enhance the robustness and generalisability of robotic inertial tracking in the field. Our scheme of choice uses continuous domain adaptation (DA) and optimal transport (OT). (3) In our evaluation, continuous OT DA outperforms a continuous adversarial DA baseline, while also showing quantifiable learning benefits over simple data augmentation. We will release our dataset to help foster future research.

arXiv:2106.14472공식 코드 · Papers with Code

Hyperbolic Busemann Learning with Ideal Prototypes

Hyperbolic space has become a popular choice of manifold for representation learning of various datatypes from tree-like structures and text to graphs. Building on the success of deep learning with prototypes in Euclidean and hyperspherical spaces, a few recent works have proposed hyperbolic prototypes for classification. Such approaches enable effective learning in low-dimensional output spaces and can exploit hierarchical relations amongst classes, but require privileged information about class labels to position the hyperbolic prototypes. In this work, we propose Hyperbolic Busemann Learning. The main idea behind our approach is to position prototypes on the ideal boundary of the Poincar\'e ball, which does not require prior label knowledge. To be able to compute proximities to ideal prototypes, we introduce the penalised Busemann loss. We provide theory supporting the use of ideal prototypes and the proposed loss by proving its equivalence to logistic regression in the one-dimensional case. Empirically, we show that our approach provides a natural interpretation of classification confidence, while outperforming recent hyperspherical and hyperbolic prototype approaches.

arXiv:2106.12075

A Hybrid Genetic-Fuzzy Controller for a 14-inches Astronomical Telescope Tracking

The performance of on telescope depend strongly on its operating conditions. During pointing the telescope can move at a relatively high velocity, and the system can tolerate trajectory position errors higher than during tracking. On the contrary, during tracking Alt-Az telescopes generally move slower but still in a large dynamic range. In this case, the position errors must be as close to zero as possible. Tracking is one of the essential factors that affect the quality of astronomical observations. In this paper, a hybrid Genetic-Fuzzy approach to control the movement of a two-link direct-drive Celestron telescope is introduced. The proposed controller uses the Genetic algorithm (GA) for optimizing a fuzzy logic controller (FLC) to improve the tracking of the 14-inches Celestron telescope of the Kottamia Astronomical Observatory (KAO). The fuzzy logic input is a vector of the position error and its rate of change, and the output is torque. The GA objective function used here is the Integral Time Absolute Error (ITAE). The proposed method is compared with a conventional Proportional-Differential (PD) controller, an optimized PD controller with a GA, and a Fuzzy controller. The results show the effectiveness of the proposed controller to improve the dynamic response of the overall system.

arXiv:2106.11537

Information of income position and its impact on perceived tax burden and preference for redistribution: An Internet Survey Experiment

A customized internet survey experiment is conducted in Japan to examine how individuals' relative income position influences preferences for income redistribution and individual perceptions regarding income tax burden. I first asked respondents about their perceived income position in their country and their preferences for redistribution and perceived tax burden. In the follow-up survey for the treatment group, I provided information on their true income position and asked the same questions as in the first survey. For the control group, I did not provide their true income position and asked the same questions. I gathered a large sample that comprised observations of the treatment group (4,682) and the control group (2,268). The key findings suggest that after being informed of individuals' real income position, (1) individuals who thought their income position was higher than the true one perceived their tax burden to be larger, (2) individuals' preference for redistribution hardly changes, and (3) irreciprocal individuals perceive their tax burden to be larger and are more likely to prefer redistribution. However, the share of irreciprocal ones is small. This leads Japan to be a non-welfare state.