Recently end-to-end scene text spotting has become a popular research topic due to its advantages of global optimization and high maintainability in real applications. Most methods attempt to develop various region of interest (RoI) operations to concatenate the detection part and the sequence recognition part into a two-stage text spotting framework. However, in such framework, the recognition part is highly sensitive to the detected results (e.g.), the compactness of text contours). To address this problem, in this paper, we propose a novel Mask AttentioN Guided One-stage text spotting framework named MANGO, in which character sequences can be directly recognized without RoI operation. Concretely, a position-aware mask attention module is developed to generate attention weights on each text instance and its characters. It allows different text instances in an image to be allocated on different feature map channels which are further grouped as a batch of instance features. Finally, a lightweight sequence decoder is applied to generate the character sequences. It is worth noting that MANGO inherently adapts to arbitrary-shaped text spotting and can be trained end-to-end with only coarse position information (e.g.), rectangular bounding box) and text annotations. Experimental results show that the proposed method achieves competitive and even new state-of-the-art performance on both regular and irregular text spotting benchmarks, i.e., ICDAR 2013, ICDAR 2015, Total-Text, and SCUT-CTW1500.
Detecting object skeletons in natural images presents challenging, due to varied object scales, the complexity of backgrounds and various noises. The skeleton is a highly compressing shape representation, which can bring some essential advantages but cause the difficulties of detection. This skeleton line occupies a rare proportion of an image and is overly sensitive to spatial position. Inspired by these issues, we propose the ProMask, which is a novel skeleton detection model. The ProMask includes the probability mask and vector router. The skeleton probability mask representation explicitly encodes skeletons with segmentation signals, which can provide more supervised information to learn and pay more attention to ground-truth skeleton pixels. Moreover, the vector router module possesses two sets of orthogonal basis vectors in a two-dimensional space, which can dynamically adjust the predicted skeleton position. We evaluate our method on the well-known skeleton datasets, realizing the better performance than state-of-the-art approaches. Especially, ProMask significantly outperforms the competitive DeepFlux by 6.2% on the challenging SYM-PASCAL dataset. We consider that our proposed skeleton probability mask could serve as a solid baseline for future skeleton detection, since it is very effective and it requires about 10 lines of code.
Photoacoustic (PA) computed tomography (PACT) reconstructs the initial pressure distribution from raw PA signals. The standard reconstruction of medical image could cause the artifacts due to interferences or ill-posed setup. Recently, deep learning has been used to reconstruct the PA image with ill-posed conditions. Most works remove the artifacts from image domain, and compensate the limited-view from dataset. In this paper, we propose a jointed feature fusion framework (JEFF-Net) based on deep learning to reconstruct the PA image using limited-view data. The cross-domain features from limited-view position-wise data and the reconstructed image are fused by a backtracked supervision. Specifically, our results could generate superior performance, whose artifacts are drastically reduced in the output compared to ground-truth (full-view reconstructed result). In this paper, a quarter position-wise data (32 channels) is fed into model, which outputs another 3-quarters-view data (96 channels). Moreover, two novel losses are designed to restrain the artifacts by sufficiently manipulating superposed data. The numerical and in-vivo results have demonstrated the superior performance of our method to reconstruct the full-view image without artifacts. Finally, quantitative evaluations show that our proposed method outperformed the ground-truth in some metrics.
We present a hybrid signaling approach to position energy-neutral devices. Our method combines the best of two worlds: instant RF signals for both communication and energy transfer, and slower propagating acoustic waves for accurate distance measurements. We introduce advanced energy loading of an 'E-buffer' through beamformed RF signals. The scientific contribution of our approach is twofold, advancing both the positioning performance and the energy harvesting efficiency. On the one hand, we overcome current distance limitations in RF backscattering-based indoor localization. On the other hand, it enhances the energy harvesting by using the calculated position and beamforming a higher amount of directed RF energy into the mobile node. We provide a functional architecture for both sides of the system and present a proof-of-concept. Practical measurements in representative use cases show an update rate of 10 positions per hour within the regulatory constraints in the 868MHz band for distances up to 4.5 m. An energy and power model are drawn up to provide insights into the performance trade-offs.
In this paper, we design a reliable Heliquad and develop an intelligent controller to handle one actuators complete failure. Heliquad is a multi-copter similar to Quadcopter, with four actuators diagonally symmetric from the center. Each actuator has two control inputs; the first input changes the propeller blades collective pitch (also called variable pitch), and the other input changes the rotation speed. For reliable operation and high torque characteristic requirement for yaw control, a cambered airfoil is used to design propeller blades. A neural network-based control allocation is designed to provide complete control authority even under a complete loss of one actuator. Nonlinear quaternion based outer loop position control, with proportional-derivative inner loop for attitude control and neural network-based control allocation is used in controller design. The proposed controller and Heliquad designs performance is evaluated using a software-in-loop simulation to track the position reference command under failure. The results clearly indicate that the Heliquad with an intelligent controller provides necessary tracking performance even under a complete loss of one actuator.
One of the main drawbacks of the well-known Direct Position Determination (DPD) method is the requirement that raw signal data be transferred to a common processor. It would therefore be of high practical value if DPD$-$or a modified version thereof$-$could be successfully applied to a coarsely quantized version of the raw data, thus alleviating the requirements on the communication links between the different base stations. Motivated by the above, and inspired by recent work in the rejuvenated one-bit array processing field, we present One-Bit DPD: a method for direct localization based on one-bit quantized measurements. We show that despite the coarse quantization, the proposed method nonetheless yields an estimate for the unknown emitter position with appealing asymptotic properties. We further establish the underlying identifiability conditions of this model, which rely only on second-order statistics. Empirical simulation results corroborate our analytical derivations, demonstrating that much of the information regarding the unknown emitter position is preserved under this crude form of quantization.
MmWave multiple-input single-output (MISO) systems using a single-antenna receiver are regarded as promising solution for the near future, before the full-fledged 5G MIMO will be widespread. However, for MISO systems synchronization cannot be performed jointly with user localization unless two-way transmissions are used. In this paper we show that thanks to the use of a reconfigurable intelligent surface (RIS), joint localization and synchronization is possible with only downlink MISO transmissions. The direct maximum likelihood (ML) estimator for the position and clock offset is derived. To obtain a good initialization for the ML optimization, a decoupled, relaxed estimator of position and delays is also devised, which does not require knowledge of the clock offset. Results show that the proposed approach attains the Cram\'er-Rao lower bound even for moderate values of the system parameters.
Humans can covertly track the position of an object, even if the object is temporarily occluded. What are the neural mechanisms underlying our capacity to track moving objects when there is no physical stimulus for the brain to track? One possibility is that the brain 'fills-in' information about imagined objects using internally generated representations similar to those generated by feed-forward perceptual mechanisms. Alternatively, the brain might deploy a higher order mechanism, for example using an object tracking model that integrates visual signals and motion dynamics. In the present study, we used EEG and time-resolved multivariate pattern analyses to investigate the spatial processing of visible and imagined objects. Participants tracked an object that moved in discrete steps around fixation, occupying six consecutive locations. They were asked to imagine that the object continued on the same trajectory after it disappeared and move their attention to the corresponding positions. Time-resolved decoding of EEG data revealed that the location of the visible stimuli could be decoded shortly after image onset, consistent with early retinotopic visual processes. For processing of unseen/imagined positions, the patterns of neural activity resembled stimulus-driven mid-level visual processes, but were detected earlier than perceptual mechanisms, implicating an anticipatory and more variable tracking mechanism. Encoding models revealed that spatial representations were much weaker for imagined than visible stimuli. Monitoring the position of imagined objects thus utilises similar perceptual and attentional processes as monitoring objects that are actually present, but with different temporal dynamics. These results indicate that internally generated representations rely on top-down processes, and their timing is influenced by the predictability of the stimulus.
The aim of this work is to control the longitudinal position of an autonomous vehicle with an internal combustion engine. The powertrain has an inherent dead-time characteristic and constraints on physical states apply since the vehicle is neither able to accelerate arbitrarily strong, nor to drive arbitrarily fast. A model predictive controller (MPC) is able to cope with both of the aforementioned system properties. MPC heavily relies on a model and therefore a strategy on how to obtain multiple linear state space prediction models of the nonlinear system via input/output data system identification from acceleration data is given. The models are identified in different regions of the vehicle dynamics in order to obtain more accurate predictions. The still remaining plant-model mismatch can be expressed as an additive disturbance which can be handled through robust control theory. Therefore modifications to the models for applying robust MPC tracking control theory are described. Then a controller which guarantees robust constraint satisfaction and recursive feasibility is designed. As a next step, modifications to apply the controller on multiple models are discussed. In this context, a model switching strategy is provided and theoretical and computational limitations are pointed out. Lastly, simulation results are presented and discussed, including computational load when switching between systems.
We propose an indoor localization algorithm for visible light systems by considering effects of non-line-of-sight (NLOS) propagation. The proposed algorithm, named database assisted nonlinear least squares (DA-NLS), utilizes ideas from both the classical NLS algorithm and the fingerprinting algorithm to achieve accurate and robust localization performance in NLOS environments. In particular, a database is used to learn NLOS effects, and then an NLS algorithm is employed to estimate the position. The performance of the proposed algorithm is compared against that of the fingerprinting and NLS algorithms.
In this paper, we propose a novel visible light communication (VLC) assisted Perspective-fourLine algorithm (V-P4L) for practical indoor localization. The basic idea of V-P4L is to joint VLC and computer vision to achieve high accuracy regardless of LED height differences. In particular, we first exploit the space-domain information to estimate the orientation and coordinate of a single rectangular LED luminaire in the camera coordinate system based on the plane and solid geometry. Then, based on the time-domain information transmitted by VLC and the estimated luminaire information, V-P4L can estimate the position and pose of the camera by the single-view geometry theory and the linear least square (LLS) method. To further mitigate the effect of height differences among LEDs on localization accuracy, we then propose a correction algorithm of V-P4L based on the LLS method and a simple optimization method. Due to the combination of time- and space-domain information, V-P4L only requires a single luminaire for localization without limitation on the correspondences between the features and their projections. Simulation results show that for V-P4L the position error is always less than 15 cm and the orientation error is always less than 3{\deg} using popular indoor luminaires.
The process of simultaneously mapping the environment in three dimensional (3D) space and localizing a moving vehicle's pose (orientation and position) is termed Simultaneous Localization and Mapping (SLAM). SLAM is a core task in robotics applications. In the SLAM problem, each of the vehicle's pose and the environment are assumed to be completely unknown. This paper takes the conventional SLAM design as a basis and proposes a novel approach that ensures fast adaptation of the nonlinear observer for SLAM. Due to the fact that the true SLAM problem is nonlinear and is modeled on the Lie group of $\mathbb{SLAM}_{n}\left(3\right)$, the proposed observer for SLAM is nonlinear and modeled on $\mathbb{SLAM}_{n}\left(3\right)$. The proposed observer compensates for unknown bias attached to velocity measurements. The results of the simulation illustrate the robustness of the proposed approach.
The Back-Projection Algorithm (BPA) is a time domain matched filtering technique to form synthetic aperture radar (SAR) images. To produce high quality BPA images, precise navigation data for the radar platform must be known. Any error in position, velocity, or attitude results in improperly formed images corrupted by shifting, blurring, and distortion. This paper develops analytical expressions that characterize the relationship between navigation errors and image formation errors. These analytical expressions are verified via simulated image formation and real data image formation.
How should one construct a portfolio from multiple mean-reverting assets? Should one add an asset to portfolio even if the asset has zero mean reversion? We consider a position management problem for an agent trading multiple mean-reverting assets. We solve an optimal control problem for an agent with power utility, and present a semi-explicit solution. The nearly explicit nature of the solution allows us to study the effects of parameter mis-specification, and derive a number of properties of the optimal solution.
Designing hardware accelerators for deep neural networks (DNNs) has been much desired. Nonetheless, most of these existing accelerators are built for either convolutional neural networks (CNNs) or recurrent neural networks (RNNs). Recently, the Transformer model is replacing the RNN in the natural language processing (NLP) area. However, because of intensive matrix computations and complicated data flow being involved, the hardware design for the Transformer model has never been reported. In this paper, we propose the first hardware accelerator for two key components, i.e., the multi-head attention (MHA) ResBlock and the position-wise feed-forward network (FFN) ResBlock, which are the two most complex layers in the Transformer. Firstly, an efficient method is introduced to partition the huge matrices in the Transformer, allowing the two ResBlocks to share most of the hardware resources. Secondly, the computation flow is well designed to ensure the high hardware utilization of the systolic array, which is the biggest module in our design. Thirdly, complicated nonlinear functions are highly optimized to further reduce the hardware complexity and also the latency of the entire system. Our design is coded using hardware description language (HDL) and evaluated on a Xilinx FPGA. Compared with the implementation on GPU with the same setting, the proposed design demonstrates a speed-up of 14.6x in the MHA ResBlock, and 3.4x in the FFN ResBlock, respectively. Therefore, this work lays a good foundation for building efficient hardware accelerators for multiple Transformer networks.
In this paper, we exploit the maximum correntropy criterion (MCC) to robustify the traditional time-difference-of-arrival (TDOA) location estimator in the presence of non-line-of-sight (NLOS) propagation conditions. For the sake of statistical efficiency, the correntropy-based robust loss is imposed on the underlying time-of-arrival composition via joint estimation of the source position and onset time, instead of the TDOA counterpart generated in the postprocessing of sensor-collected timestamps. We then employ a neurodynamic optimization approach to tackle the highly nonconvex MCC formulation. Furthermore, we examine the local stability of equilibrium for the corresponding projection-type neural network model. Simulation investigations in representative NLOS propagation scenarios demonstrate that our neurodynamic robust TDOA localization solution is capable of outperforming several existing schemes in terms of positioning accuracy.
This technical communiqu\'e aims at correcting an erroneous statement (Lemma 2.4) in an earlier paper by the same authors concerning a sufficient condition of uniform observability for a Linear Time-Varying (LTV) system. In this earlier paper, the proofs of two other lemmas, about body-pose estimation from range measurements, relied on this erroneous statement. For the sake of conciseness, only a new proof of one of these lemmas is presented, the proof of the second lemma being a simpler version of it.
The paper explores the application of a continuous action space soft actor-critic (SAC) reinforcement learning model to the area of automated market-making. The reinforcement learning agent receives a simulated flow of client trades, thus accruing a position in an asset, and learns to offset this risk by either hedging at simulated "exchange" spreads or by attracting an offsetting client flow by changing offered client spreads (skewing the offered prices). The question of learning minimum spreads that compensate for the risk of taking the position is being investigated. Finally, the agent is posed with a problem of learning to hedge a blended client trade flow resulting from independent price processes (a "portfolio" position). The position penalty method is introduced to improve the convergence. An Open-AI gym-compatible hedge environment is introduced and the Open AI SAC baseline RL engine is being used as a learning baseline.
In this paper, a position and velocity estimation method for robotic manipulators which are affected by constant bounded disturbances is considered. The tracking control problem is formulated as a disturbance rejection problem, with all the unknown parameters and dynamic uncertainties lumped into disturbance. Using adaptive robust extended Kalman filter(AREKF) the movement and velocity of each joint is predicted to use in discontinuous Lyapunov-based controller structure. The parameters of the error dynamics have been validated off-line by real data. Computer simulation results given for a two degree of freedom manipulator demonstrate the efficacy of the improved Kalman Filter by comparing the performance of EKF and improved AREKF. Although it is shown that accurate trajectory tracking can be achieved by using the proposed controller.
The Astronomical Observatory of the Sergio Arboleda University (Bogota, Colombia) has as main instrument a Torus Classic Cassegrain telescope. In order to improve its precision and accuracy in the tracking of celestial objects, the Torus telescope requires to be properly automated. The motors, drivers and sensors of the telescope are modelled as a black-box systems and experimentally identified. Using the identified model several control laws are developed such as Proportional, Integral and Derivative (PID) control and State-feedback control for the velocity and position tracking and disturbance-rejection tasks.