Targets In Close Proximity Research Articles

Real-Time 3D Object Detection on Crowded Pedestrians.

In the field of autonomous driving, object detection under point clouds is indispensable for environmental perception. In order to achieve the goal of reducing blind spots in perception, many autonomous driving schemes have added low-cost blind-filling LiDAR on the side of the vehicle. Unlike point cloud target detection based on high-performance LiDAR, the blind-filling LiDARs have low vertical angular resolution and are mounted on the side of the vehicle, resulting in easily mixed point clouds of pedestrian targets in close proximity to each other. These characteristics are harmful for target detection. Currently, many research works focus on target detection under high-density LiDAR. These methods cannot effectively deal with the high sparsity of the point clouds, and the recall and detection accuracy of crowded pedestrian targets tend to be low. To overcome these problems, we propose a real-time detection model for crowded pedestrian targets, namely RTCP. To improve computational efficiency, we utilize an attention-based point sampling method to reduce the redundancy of the point clouds, then we obtain new feature tensors by the quantization of the point cloud space and neighborhood fusion in polar coordinates. In order to make it easier for the model to focus on the center position of the target, we propose an object alignment attention module (OAA) for position alignment, and we utilize an additional branch of the targets' location occupied heatmap to guide the training of the OAA module. These methods improve the model's robustness against the occlusion of crowded pedestrian targets. Finally, we evaluate the detector on KITTI, JRDB, and our own blind-filling LiDAR dataset, and our algorithm achieved the best trade-off of detection accuracy against runtime efficiency.

MO-0640 Structural cardiac changes in cardiac MRI after SBRT for targets in close proximity to the heart

MO-0640 Structural cardiac changes in cardiac MRI after SBRT for targets in close proximity to the heart

A numerical simulation of target-directed swimming for a three-link bionic fish with deep reinforcement learning

An accurate and robust feedback control system is of great importance for bionic robotic fish to perform complex tasks. This work presents a numerical study of target-directed swimming for a three-link bionic fish with a feedback control system based on deep reinforcement learning (DRL). The simulation is achieved by using a hybrid method of the DRL method and the immersed boundary–lattice Boltzmann method (IB–LBM). This framework makes use of the high computational efficiency of the IB-LBM for the generation of massive motion data needed in the DRL training. The fish is first trained to swim towards a static target from random orientation and distance. The only information available to the fish is its orientation and distance from the target. It learns an accurate and subtle control policy after the training. Then the control policy is applied to a moving target. Even though the fish encounters some situations that never happened in the training, it can choose appropriate actions to follow the target in close proximity and automatically adapt its velocity with the velocity of the target. Those simulations demonstrate the accuracy, robustness of the control method, and its ability to adapt to new situations.

A Dataset of Ground Vehicle Targets from Satellite SAR Images and Its Application to Detection and Instance Segmentation

The advent of deep learning-based algorithms has facilitated researches on target detection from synthetic aperture radar(SAR) imagery. While most of them concentrate on detection tasks for ships with open SAR ship datasets and for aircraft from SAR scenes of airports, there is relatively scarce researches on the detection of SAR ground vehicle targets where several adverse factors such as high false alarm rates, low signal-to-clutter ratios, and multiple targets in close proximity are predicted to degrade the performances. In this paper, a dataset of ground vehicle targets acquired from TerraSAR-X(TSX) satellite SAR images is presented. Then, both detection and instance segmentation are simultaneously carried out on this dataset based on the deep learning-based Mask R-CNN. Finally, this paper shows the future research directions to further improve the performances of detecting the SAR ground vehicle targets.

Multi-Target Tracking Using Windowed Fourier Single-Pixel Imaging.

The single-pixel imaging (SPI) technique enables the tracking of moving targets at a high frame rate. However, when extended to the problem of multi-target tracking, there is no effective solution using SPI yet. Thus, a multi-target tracking method using windowed Fourier single-pixel imaging (WFSI) is proposed in this paper. The WFSI technique uses a series of windowed Fourier basis patterns to illuminate the target. This method can estimate the displacements of K independently moving targets by implementing measurements and calculating windowed Fourier coefficients, which is a measurement method with low redundancy. To enhance the capability of the proposed method, we propose a joint estimation approach for multi-target displacement, which solves the problem where different targets in close proximity cannot be distinguished. Using the independent and joint estimation approaches, multi-target tracking can be implemented with WFSI. The accuracy of the proposed multi-target tracking method is verified by numerical simulation to be less than 2 pixels. The tracking effectiveness is analyzed by a video experiment. This method provides, for the first time, an effective idea of multi-target tracking using SPI.

A multi-frame track-before-detect algorithm based on root label clustering for multiple targets

In this paper, a novel multi-frame track-before-detect algorithm is proposed, which is based on root label clustering to reduce the high computational complexity arising by observation area expansion and clutter/noise density increase. A criterion of track extrapolation is used to construct state transition set, root label is marked by state transition set to obtain the distribution information of multiple targets in measurement space, then measurement plots of multi-frame are divided into several clusters, and finally multi-frame track-before-detect algorithm is implemented in each cluster. The computational complexity can be reduced by employing the proposed algorithm. Simulation results show that the proposed algorithm can accurately detect multiple targets in close proximity and reduce the number of false tracks.

Unravelling Ipsilateral Interactions Between Left Dorsal Premotor and Primary Motor Cortex: A Proof of Concept Study

Few studies have identified the intrahemispheric functional connectivity between the ipsilateral dorsal premotor cortex (PMd) and the primary motor hand area (M1hand) due to technical limitations. In this proof-of-concept study, a novel neuronavigated dsTMS set-up was employed, combining stimulation over left PMd and left M1hand using the edge of a butterfly coil and a small cooled-coil. This arrangement was warranted because coil (over)heating and inter coil distance are limiting factors when investigating connectivity between stimulation targets in close proximity and over a longer duration. The proposed set-up was designed to deal with these limitations. Specifically, the effect of four dual-site transcranial magnetic stimulation (dsTMS) protocols on twenty-eight right-handed participants (12 males) was evaluated. These protocols differed in stimulus order, interstimulus interval and current direction induced in PMd. A structural scan with electric (E-)field modeling was obtained from seven participants prior to dsTMS, demonstrating that PMd and M1hand were effectively stimulated. Results indicate that one protocol, in which a latero-medial current was induced in PMd 2.8 ms prior to stimulation over M1hand, induced a sex-mediated effect. In males, significant inhibition of motor-evoked potentials was identified, whereas females demonstrated a facilitatory effect that did not survive correction for multiple comparisons. E-field simulations revealed that the E-field induced by the coil targeting PMd was maximal in PMd, with weaker E-field strengths extending to regions beyond PMd. Summarizing, the current dsTMS set-up enabled stimulating at an inter-target distance of 35 mm without any indications of coil-overheating.

An examination of mobile phone pointing in surface mapped spatial augmented reality

We investigate mobile phone pointing in Spatial Augmented Reality (SAR), where digital content is mapped onto the surfaces of a real physical environment. Three pointing techniques are compared: raycast, viewport, and direct. A first experiment examines these techniques in a realistic five-projector SAR environment with representative targets distributed across different surfaces. Participants were permitted free movement, so variations in target occlusion and target view angle occurred naturally. A second experiment validates and further generalizes findings by strictly controlling target occlusion and view angle in a simulated SAR pointing task using an AR HMD. Overall, results show raycast is fastest for non-occluded targets, direct is most accurate, and fastest for occluded targets in close proximity, and viewport falls in between. Using the experiment data, we formulate and evaluate a new Fitts’ model combining two spatial configurations in a SAR pointing task to capture key characteristics, initial target occlusion, target view angle, and user movement. Analysis shows it is a better predictor than previous models.

Analyzing the Integrin Adhesome by In Situ Proximity Ligation Assay.

The in situ proximity ligation assay (PLA) is capable of detecting single protein events such as protein protein-interactions and posttranslational modifications (e.g., protein phosphorylation) in tissue and cell samples prepared for analysis by immunofluorescent or immunohistochemical microscopy. The targets are detected using two primary antibodies which must be from different host species. A pair of secondary antibodies (PLA probes) conjugated to complementary oligonucleotides is applied to the sample, and a signal is generated only when the two PLA probes are in close proximity by their binding to the two primary antibodies that have bound to their targets in close proximity. The signal from each pair of PLA probes is visualized as an individual fluorescent spot. These PLA signals can be quantified (counted) using image analysis software (ImageJ), and also assigned to a specific subcellular location based on microscopy image overlays. In principle, in situ PLA offers a relatively simple and sensitive technique to analyze interactions among any proteins for which suitable antibodies are available. Integrin-mediated focal adhesions (FAs) are large multiprotein complexes consisting of more than 150 proteins, also known as the integrin adhesome, which link the extracellular matrix (ECM) to the actin cytoskeleton and regulate the functioning of mechanosignaling pathways. The in situ PLA approach is well suited for examining the spatiotemporal aspects of protein posttranslational modifications and protein interactions occurring in dynamic multiprotein complexes such as integrin mediated focal adhesions.

An improved merging method for Gaussian mixture probability hypothesis density filter

The random finite set (RFS)-based Gaussian mixture probability hypothesis density (GM-PHD) filter is a promising and efficient suboptimal approximation for the multi-target Bayes filter. However, the GM-PHD filter is unable to track nearby targets caused by the improper position distribution of target-originated measurements. Aiming at the problem, a multi-target GM-PHD filter with an improved component merging method is proposed. Based on a proposed adaptive threshold-based component similarity measure scheme, the improved component merging method is able to avoid incorrect fusion of the components of targets in close proximity and optimize the target components within the target posterior intensity. Experimental results illustrate that the proposed algorithm not only can achieve better estimation accuracy in terms of the target states and its number but also has high computation efficiency when compared against the related GM-PHD-based filters.

Accurate Ultra-Wideband Array Radar Imaging Using Four-Dimensional Unitary ESPRIT

Adaptive signal processing techniques can estimate the directions of arrival (DOAs), ranges, and radial velocities of radar targets with high resolution when applied in the space, frequency, and time domains, respectively. However, the performance of these techniques is limited when they are applied separately in each domain. Recently, adaptive signal processing algorithms based on high-dimensional signal subspaces have been studied extensively. We apply a four-dimensional unitary estimation of signal parameters via rotational invariant techniques (ESPRIT) algorithm for the simultaneous estimation of the DOAs, ranges, and Doppler velocities of multiple targets in an ultra-wideband radar imaging setting and show the effectiveness of the high-dimensional ESPRIT for the near-field imaging of distributed moving targets. The four-dimensional ESPRIT algorithm is demonstrated to be able to separate targets moving in close proximity, whereas the two- and three-dimensional ESPRIT algorithms fail to separate these targets accurately because of their limited resolutions. The application of the high-dimensional ESPRIT to near-field radar imaging covers a wide range of applications that require the measurement of multiple moving targets in close proximity. An example of such an application is radar-based human monitoring. Therefore, the superior resolution of the high-dimensional ESPRIT has the potential to improve the performance of various real-world security and healthcare systems.

Multiscale Rotated Bounding Box-Based Deep Learning Method for Detecting Ship Targets in Remote Sensing Images.

Since remote sensing images are captured from the top of the target, such as from a satellite or plane platform, ship targets can be presented at any orientation. When detecting ship targets using horizontal bounding boxes, there will be background clutter in the box. This clutter makes it harder to detect the ship and find its precise location, especially when the targets are in close proximity or staying close to the shore. To solve these problems, this paper proposes a deep learning algorithm using a multiscale rotated bounding box to detect the ship target in a complex background and obtain the location and orientation information of the ship. When labeling the oriented targets, we use the five-parameter method to ensure that the box shape is maintained rectangular. The algorithm uses a pretrained deep network to extract features and produces two divided flow paths to output the result. One flow path predicts the target class, while the other predicts the location and angle information. In the training stage, we match the prior multiscale rotated bounding boxes to the ground-truth bounding boxes to obtain the positive sample information and use it to train the deep learning model. When matching the rotated bounding boxes, we narrow down the selection scope to reduce the amount of calculation. In the testing stage, we use the trained model to predict and obtain the final result after comparing with the score threshold and nonmaximum suppression post-processing. Experiments conducted on a remote sensing dataset show that the algorithm is robust in detecting ship targets under complex conditions, such as wave clutter background, target in close proximity, ship close to the shore, and multiscale varieties. Compared to other algorithms, our algorithm not only exhibits better performance in ship detection but also obtains the precise location and orientation information of the ship.

Pulsed laser deposition of plasmonic nanostructured gold on flexible transparent polymers at atmospheric pressure

In this paper, we outline a novel technique for the deposition of nanostructured thin films utilizing a modified form of pulsed laser deposition (PLD). We demonstrate confined atmospheric PLD (CAP) for the deposition of gold on cyclic olefin polymer substrates. The deposition process is a simplified form of conventional PLD, with deposition conducted under atmospheric conditions and the substrate and target in close proximity. It was found that this confinement results in the deposition of nanostructured thin films on the substrate. Infrared spectroscopy showed no significant change of polymer surface chemistry as a result of the deposition process, and optical spectroscopy revealed plasmonic behavior of the resulting thin film. The effect of laser fluence on the deposition process was also examined with more uniform films deposited at higher fluences.

Multi‐target trackers using cubature Kalman filter for Doppler radar tracking in clutter

A major feature of the Gaussian mixture probability hypothesis density (GM-PHD) filter is that it does not require any measurement-to-track association to complete its update step. This, according to the authors, should constitute significant advantage over conventional data-association based methods, especially in presence of high false-alarm rate, frequent miss-detections and targets in close proximity. To test this hypothesis, a multi-target tracking (MTT) problem using Doppler radar is considered, where the performance of GM-PHD algorithm is compared against six data-association based MTT filters in aforementioned adverse tracking conditions. To handle the non-linearity due to Doppler, cubature Kalman filter (CKF) is used in the framework of all MTT algorithms. Detailed mathematical framework of a new non-linear variant of GM-PHD using CKF has been derived using fundamental principles of non-linear Bayesian filtering. It is named as CK-GM-PHD. CK-GM-PHD is formulated using approximated Gaussian mixture assumption and follows track-oriented approach. Cubature integration method is used to numerically compute mean and covariance of components in the Gaussian mixture. Simulation results support the hypothesis by revealing substantial performance improvement of CK-GM-PHD algorithm over conventional data-association based approaches while tested in moderate to heavy clutter rate with lower detection probability and closely spaced target scenarios.

Maintaining Track Continuity for Extended Targets Using Gaussian-Mixture Probability Hypothesis Density Filter

A multiextended-target tracker based on the extended target Gaussian-mixture probability hypothesis density (ET-GMPHD) filter, which can provide the tracks of the extended targets, is proposed to maintain the track continuity for the extended targets. To identify the extended targets, each individual Gaussian term of the mixture representing the posterior intensity function will be assigned a label, which is evolved through time. Then a track management scheme, including track initiation, track confirmation, track propagation, and termination, is developed to form the tracks for the extended targets. Furthermore, to improve the performance of the extended target tracker we also propose a mixture partitioning algorithm for resolving the identities of the extended targets in close proximity. The simulation results show that our proposed tracker achieves the less error of the position estimates and decreases the probability of incorrect label assignments from 0.6 to 0.25.

NT-30 * COMPLICATIONS OF LASER INTERSTITIAL THERMAL THERAPY IN PATIENTS WITH BRAIN TUMORS

OBJECTIVE: Laser interstitial thermal therapy (LITT) has been described as a treatment for patients with brain tumors refractory to other treatments, with little emphasis on the risks of the procedure. This report focuses on complications of LITT in patients with tumors of the central nervous system. METHODS: Laser probes were inserted in the operating room using low-field intraoperative MRI (iMRI) guidance (5 patients), frameless stereotaxy (FS) in 2, and frame-based placement in 3. LITT itself was done in a 1.5 Tesla diagnostic MRI. Diagnoses included metastatic tumors in 7 patients, glioma in 2, and pituitary tumor in 1. One patient had a filum terminale ependymoma (with laser placement in the diagnostic MRI). RESULTS: In 1 patient in whom FS was used for image guidance, diagnostic MRI showed the laser fiber to be misplaced. Three complications resulted from the LITT itself. One patient with a glioblastoma of the vermis and fourth ventricle had bilateral palsies of cranial nerves 6 and 7 after treatment. A patient with a recurrent metastatic tumor after stereotactic radiosurgery had worsened left hemiparesis. These two patients were treated with steroids, with partial improvement of their deficits. The patient with the filum terminale ependymoma developed a paraparesis the day after treatment; MRI showed expansion of the intraspinal mass with resolution of enhancement, consistent with tumor ablation. She required a laminectomy and removal of the mass, with a combination of tumor and necrosis on histological examination. CONCLUSIONS: Complications of LITT can result from laser misplacement and from the laser treatment itself. Use of a stereotactic frame can limit the risk of inaccurate placement. Avoiding complications of the LITT itself can be done by undertreating the margin of a target in close proximity to critical structures. Intraspinal LITT should be used with great caution.

Engineering self-contained DNA circuit for proximity recognition and localized signal amplification of target biomolecules.

Biomolecular interactions have important cellular implications, however, a simple method for the sensing of such proximal events is lacking in the current molecular toolbox. We designed a dynamic DNA circuit capable of recognizing targets in close proximity to initiate a pre-programmed signal transduction process resulting in localized signal amplification. The entire circuit was engineered to be self-contained, i.e. it can self-assemble onto individual target molecules autonomously and form localized signal with minimal cross-talk. α-thrombin was used as a model protein to evaluate the performance of the individual modules and the overall circuit for proximity interaction under physiologically relevant buffer condition. The circuit achieved good selectivity in presence of non-specific protein and interfering serum matrix and successfully detected for physiologically relevant α-thrombin concentration (50 nM–5 μM) in a single mixing step without any further washing. The formation of localized signal at the interaction site can be enhanced kinetically through the control of temperature and probe concentration. This work provides a basic general framework from which other circuit modules can be adapted for the sensing of other biomolecular or cellular interaction of interest.

Collaborative penalized Gaussian mixture PHD tracker for close target tracking

Gaussian mixture probability hypothesis density (GM-PHD) recursion is a promising, computationally tractable implementation for the probability hypothesis density (PHD) filter. The competitive GM-PHD (CGM-PHD) and penalized GM-PHD (PGM-PHD) filters employ renormalization schemes to refine the weights assigned to each target and improve the estimation performance of the GM-PHD filter for closely spaced targets. However, these methods do not provide target trajectories over time, and the problem of wrongly identifying close targets is not still solved for the GM-PHD tracker. In this paper, we propose a collaborative penalized scheme to overcome the drawbacks of the GM-PHD tracker using the track label of each Gaussian component in the GM-PHD recursion. The simulation results show that the collaborative penalized GM-PHD (CPGM-PHD) tracker not only improves the estimation accuracy of the number of target and states but also provides the correct identities of targets in close proximity.

Nucleolin Mediates MicroRNA-directed CSF-1 mRNA Deadenylation but Increases Translation of CSF-1 mRNA

CSF-1 mRNA 3'UTR contains multiple unique motifs, including a common microRNA (miRNA) target in close proximity to a noncanonical G-quadruplex and AU-rich elements (AREs). Using a luciferase reporter system fused to CSF-1 mRNA 3'UTR, disruption of the miRNA target region, G-quadruplex, and AREs together dramatically increased reporter RNA levels, suggesting important roles for these cis-acting regulatory elements in the down-regulation of CSF-1 mRNA. We find that nucleolin, which binds both G-quadruplex and AREs, enhances deadenylation of CSF-1 mRNA, promoting CSF-1 mRNA decay, while having the capacity to increase translation of CSF-1 mRNA. Through interaction with the CSF-1 3'UTR miRNA common target, we find that miR-130a and miR-301a inhibit CSF-1 expression by enhancing mRNA decay. Silencing of nucleolin prevents the miRNA-directed mRNA decay, indicating a requirement for nucleolin in miRNA activity on CSF-1 mRNA. Downstream effects followed by miR-130a and miR-301a inhibition of directed cellular motility of ovarian cancer cells were found to be dependent on nucleolin. The paradoxical effects of nucleolin on miRNA-directed CSF-1 mRNA deadenylation and on translational activation were explored further. The nucleolin protein contains four acidic stretches, four RNA recognition motifs (RRMs), and nine RGG repeats. All three domains in nucleolin regulate CSF-1 mRNA and protein levels. RRMs increase CSF-1 mRNA, whereas the acidic and RGG domains decrease CSF-1 protein levels. This suggests that nucleolin has the capacity to differentially regulate both CSF-1 RNA and protein levels. Our finding that nucleolin interacts with Ago2 indirectly via RNA and with poly(A)-binding protein C (PABPC) directly suggests a nucleolin-Ago2-PABPC complex formation on mRNA. This complex is in keeping with our suggestion that nucleolin may work with PABPC as a double-edged sword on both mRNA deadenylation and translational activation. Our findings underscore the complexity of nucleolin's actions on CSF-1 mRNA and describe the dependence of miR-130a- and miR-301a-directed CSF-1 mRNA decay and inhibition of ovarian cancer cell motility on nucleolin.

Respiratory Monitoring and Clutter Rejection Using a CW Doppler Radar With Passive RF Tags

A Doppler radar system was built to detect the motion of modulated backscatter data from a passive tag, while rejecting the unmodulated backscatter from other elements. A harmonic tag was designed and fabricated to show the feasibility of sensing cardiopulmonary related Doppler shift in the received second harmonic signal. A fundamental frequency of 2.45 GHz was transmitted at a target and a quadrature homodyne receiver was used to sense the received second harmonic signal. The setup was used to first successfully isolate periodic motion from a tag on a programmable mechanical target in close proximity to other moving objects and to sense respiratory motion of a human subject. Respiratory data obtained from the tag was compared to a reference. The harmonic radar setup was then successfully used to isolate the respiration of a tagged human subject from an untagged large scattering object. The response of the harmonic radar was also compared to a 2.45 GHz Doppler radar for some experiment scenarios that clearly indicate the utility and potential for such systems.