Real-time Head Research Articles

Adaptive positioning of large-scale aircraft components using a simplified image-based visual servoing via online measurement of ball-head center

In digital alignment systems, ensuring fast and accurate alignment between the ball head on large aircraft components and the ball socket of the positioner is the key to aircraft assembly efficiency and quality. Image-based visual servoing is one of the most advanced techniques with high efficiency and positioning accuracy, but it does not account for measuring the ball-head position. Moreover, variable target positions, similar feature interference, and poor lighting conditions pose great challenges to the fast and robust visual measurement of ball-head position. Hence, an online visual measurement method is introduced to measure the absolute position of the ball head in real time. An adaptive region of interest extraction algorithm and an improved edge following method are introduced to extract the elliptical contour from the ball-head image, and ellipse detection is achieved through a simple triplet selection algorithm. A geometric model based on the perspective projection of the sphere is established for estimating the ball-head center. Then, the simplified image-based visual servoing method is developed by constantizing the image Jacobian matrix. Next, an adaptive positioning method of large-scale aircraft components with multiple ball heads is proposed from a holistic perspective. Finally, experiments show that the final positioning accuracy of each ball head in the X/Y plane is less than ± 0.05 mm, with an efficiency almost ten times that of the conventional method. After adaptive adjustment, the position and angle errors of the simulated component are reduced within 0.6 mm and 0.3°, respectively.

Laboratory Dangerous Operation Behavior Detection System Based on Deep Learning Algorithm

Aiming at the problem that dangerous operation behaviors in the laboratory is difficult to identify by monitoring the video. An algorithm of dangerous operation behavior detection in multi-task laboratory based on improved YOLOv5 structure is proposed. Firstly, the algorithm enhances, adaptively scales, and adaptively anchors box computing on the input of YOLO network. Then convolution operation is carried out to strengthen the ability of network feature fusion. Finally, the GIoU_Loss function is used at the output to optimize the network parameters and accelerate the convergence of the model. The experimental results show that the algorithm performs well in real-time head localization, head segmentation, and population regression, with significant innovation and superiority. Compared with traditional methods, this algorithm has better accuracy and real-time performance and can more effectively achieve human operation behaviors detection in laboratory application environments.

Heads Up! A Biomechanical Pilot Investigation of Soccer Heading Using Instrumented Mouthguards (iMGs)

Soccer players purposefully head the ball, raising concerns about reduced tolerance to concussion and potential long-term brain health. By combining qualitative video analysis with custom-fit instrumented mouthguards (iMGs), we aimed to categorize header kinematics (peak linear acceleration (PLA) and peak angular acceleration (PAA)) by header type and ball delivery method. iMGs were fitted to 10 male collegiate players for twelve matches. A total of 133 headers were verified and contextualized via video review. The most common header type (38.7%), as well as the preceding ball delivery method (47.4%), was found to be a pass. Approximately one-quarter of header impacts (27.0%) occurred below 10 g. For header type, there were no significant differences in kinematics, with shot attempts having the highest median PLA and PAA. For ball delivery methods, goal kicks had significantly greater PAA than long balls and pass attempts. The current study highlights the utility of qualitative video analysis in combination with real-time head kinematic data from iMGs to understand the mechanism and severity of header impacts. The pilot findings indicate that high-speed ball delivery methods result in higher head kinematics and should be a focus of future mitigation strategies.

A Novel Zernike Moment-Based Real-Time Head Pose and Gaze Estimation Framework for Accuracy-Sensitive Applications.

A real-time head pose and gaze estimation (HPGE) algorithm has excellent potential for technological advancements either in human-machine or human-robot interactions. For example, in high-accuracy advent applications such as Driver's Assistance System (DAS), HPGE plays a crucial role in omitting accidents and road hazards. In this paper, the authors propose a new hybrid framework for improved estimation by combining both the appearance and geometric-based conventional methods to extract local and global features. Therefore, the Zernike moments algorithm has been prominent in extracting rotation, scale, and illumination invariant features. Later, conventional discriminant algorithms were used to classify the head poses and gaze direction. Furthermore, the experiments were performed on standard datasets and real-time images to analyze the accuracy of the proposed algorithm. As a result, the proposed framework has immediately estimated the range of direction changes under different illumination conditions. We obtained an accuracy of ~85%; the average response time was 21.52 and 7.483 ms for estimating head poses and gaze, respectively, independent of illumination, background, and occlusion. The proposed method is promising for future developments of a robust system that is invariant even to blurring conditions and thus reaching much more significant performance enhancement.

Real-time and Recursive Estimators for Functional MRI Quality Assessment.

Real-time quality assessment (rtQA) of functional magnetic resonance imaging (fMRI) based on blood oxygen level-dependent (BOLD) signal changes is critical for neuroimaging research and clinical applications. The losses of BOLD sensitivity because of different types of technical and physiological noise remain major sources of fMRI artifacts. Due to difficulty of subjective visual perception of image distortions during data acquisitions, a comprehensive automatic rtQA is needed. To facilitate rapid rtQA of fMRI data, we applied real-time and recursive quality assessment methods to whole-brain fMRI volumes, as well as time-series of target brain areas and resting-state networks. We estimated recursive temporal signal-to-noise ratio (rtSNR) and contrast-to-noise ratio (rtCNR), and real-time head motion parameters by a framewise rigid-body transformation (translations and rotations) using the conventional current to template volume registration. In addition, we derived real-time framewise (FD) and micro (MD) displacements based on head motion parameters and evaluated the temporal derivative of root mean squared variance over voxels (DVARS). For monitoring time-series of target regions and networks, we estimated the number of spikes and amount of filtered noise by means of a modified Kalman filter. Finally, we applied the incremental general linear modeling (GLM) to evaluate real-time contributions of nuisance regressors (linear trend and head motion). Proposed rtQA was demonstrated in real-time fMRI neurofeedback runs without and with excessive head motion and real-time simulations of neurofeedback and resting-state fMRI data. The rtQA was implemented as an extension of the open-source OpenNFT software written in Python, MATLAB and C++ for neurofeedback, task-based, and resting-state paradigms. We also developed a general Python library to unify real-time fMRI data processing and neurofeedback applications. Flexible estimation and visualization of rtQA facilitates efficient rtQA of fMRI data and helps the robustness of fMRI acquisitions by means of substantiating decisions about the necessity of the interruption and re-start of the experiment and increasing the confidence in neural estimates.

TRFH: towards real-time face detection and head pose estimation

Nowadays, face detection and head pose estimation have a lot of application such as face recognition, aiding in gaze estimation and modeling attention. For these two tasks, it is usually to design two different models. However, the head pose estimation model often depends on the region of interest (ROI) detected in advance, which means that a serial face detector is needed. Even the lightest face detector will slow down the whole forward inference time and cannot achieve real-time performance when detecting the head pose of multiple people. We can see that both face detection and head pose estimation need face features, so a shared face feature map can be used between them. In this paper, a multi-task learning model is proposed that can solve both problems simultaneously. We directly detect the location of the center point of the bounding box of face; at this location, we calculate the size of the bounding box of face and the head attitude. We evaluate our model’s performance on the AFLW. The proposed model has great competitiveness with the multi-stage face attribute analysis model, and our model can achieve real-time performance.

Dynamic material deposition control for large-scale additive manufacturing

Large-scale additive manufacturing involves fabricating parts by joint printing of materials layer upon layer. The product quality and process efficiency are yet to be addressed to guarantee the process viability in practice. The print surface temperature has a significant impact on both of these elements and can be controlled by properly scheduling the material depositions on the surface. The thermal infrared images captured in real-time are processed, and the extracted thermal profiles are translated into a nonlinear profile model describing the heat dissipation on the surface. A real-time layer time control model is formulated to determine the best time to print the next layer. Furthermore, exploiting the maneuverability characteristics of the printer head while considering its mechanical constraints, a real-time printer head speed control model is formulated as a nonlinear mixed-integer program. Following the deterministic finite-state optimal control and shortest path problem paradigm, a novel algorithm is developed to decide the optimal printing speed trajectory for each layer. The proposed approach was tested by two case studies, including a thin wall specimen and a car lower chassis. The results showed that the method can capture the thermodynamics of the process and achieve simultaneous improvement in both quality and efficiency.

Sound Field Projection System using Optical See-Through Head Mounted Display

There are various ways to grasp the spatial and temporal structures of sound field. Sound field visualization is an effective technique to understand spatial sound information. For example, acoustical holography, optical methods, and beam-forming have been proposed and studied. In recent years, augmented reality (AR) technology has rapidly developed and is now more familiar. Many sensors, display devices, and ICT technologies have been implemented in AR equipment, which enable interaction between real and virtual worlds. In this paper, we propose an AR display system, which displays the results obtained by the beam-forming method. The system consists of 16ch microphone array, real-time sound field visualization system and optical see-through head mounted display (OST-HMD). Real-time sound field visualization system analyses sound signals recorded by 16ch microphone array by beam-forming method. Processed sound pressures data are sent to OST-HMD by using transmission control protocol (TCP), and colormap is projected on real world. Settings property of real-time sound field visualization system can be changed by using virtual user interface (UI) and TCP. In addition, multi-users can experience the system by sharing sound pressures and settings property data. Using this system, users wearing OST-HMD can observe sound field information intuitively.

HEAD MOTION DETECTION

The proposed system is a system that can be used for surveillance and monitoring applications. The development of an efficient real-time video head detection system is motivated by its potential for deployment in the areas where security is the main concern. The proposed system presents a platform for real-time video head detection and subsequent generation of an alarm condition as soon as the human head is detected. The prototype consists of a platform mounted with a camera that provides continuous feedback of the online exam environment.

Recognition of cutting rock hardness by hydraulic cylinder pressure

The online identification of rock hardness is the basis to adjust the rotation velocity and swing velocity of cutting head in real time. To effectively identify the hardness of rock around roadway under different cutting conditions, a novel method based on the hydraulic cylinder pressure signal is proposed. Here, we analyze the changing rule of cutting head load under different rock hardness through dynamic simulation, and the transfer characteristics between cutting dynamic load and the hydraulic cylinder pressure. On this basis, a function model between the hydraulic cylinder pressure and the protodyakonov scale of rock hardness is constructed to identify the rock hardness. The average recognition accuracy of the heavy-duty roadheader in the test is 92.01%. The experimental results indicate that the model can effectively identify the rock hardness under different cutting conditions. This study can be used for online identification of rock hardness and provide reliable basis for the automatic control of roadheader.

Profile of a patient with non-ST segment elevation myocardial infarction in actual clinical practice

Aim. To describe profile of a modern portrait with non-ST-segment elevation myocardial infarction (non-STEMI) through a comprehensive analysis of the Emergency Cardiology Unit (ECU) practice, which discharge a function of a regional vascular centre.Material and methods. To describe the non-STEMI trends of the last decade, we analysed the annual reports on ECU work. The main analysis included patients with a documented non-STEMI treated in 2019 (n=221). We used information from the department database. A Microsoft Excel software was used to create the database. The base has been filled in by the ECU head in real time since 2009. Statistical data processing was performed using the Statistica 10,0 software package. The methods of descriptive statistics and Yates-corrected chi-square test were used.Results. The following clinical and demographic trends of the last decade were revealed: an increase in the number of patients with non-STEMI, proportion of male patients, mean age of patients, proportion of patients with MI with non-obstructive coronary artery disease; no decrease in in-hospital mortality, despite the introduction of modern guidelines, pharmacological and invasive treatment of non-STEMI. In 2019, the proportion of male patients and patients 75 years and older was 62,4% and 32%, respectively. The mean age of patients was 64,6±13,0 years. Clopidogrel was the predominant P2Y12 receptor blockers (56,1%). A total of 176 patients (79,6%) underwent the invasive procedures. Endovascular myocardial revascularization was performed in 97 patients (43,9%), while in the group over 75 years old — in 16 (7%) patients. The leading causes for absence of myocardial revascularization were chronic kidney disease (4,6%), severe coronary artery disease (6,3%), “borderline” (50-60%) coronary artery stenosis. The overall in-hospital mortality rate was 9,0%, while in the group of patients over 75 years old — 19,7%. Mortality rates did not differ in patients with and without myocardial revascularization (p=0,2). However, the incidence of pulmonary oedema was higher in the conservative treatment group (p=0,04).Conclusion. Treatment of patients 75 years and older remains the main barrier in management of patients with non-STEMI. We observe the treatment-risk paradox, which consists in choosing a less aggressive treatment strategy in the group of the most high-risk patients. Other relevant aspects in the management of non-STEMI patients are the selection of a method for myocardial revascularization in multivessel coronary artery disease, assessment of the hemodynamic significance of coronary artery stenosis, and patients with non-obstructive coronary artery disease.

Real-Time Driver’s Focus of Attention Extraction and Prediction using Deep Learning

Driving is one of the most common activities in our modern lives. Every day, millions drive to and from their schools or workplaces. Even though this activity seems simple and everyone knows how to drive on roads, it actually requires drivers’ complete attention to keep their eyes on the road and surrounding cars for safe driving. However, most of the research focused on either keeping improving the configurations of active safety systems with high-cost components like Lidar, night vision cameras, and radar sensor array, or finding the optimal way of fusing and interpreting sensor information without consid-ering the impact of drivers’ continuous attention and focus. We notice that effective safety technologies and systems are greatly affected by drivers’ attention and focus. In this paper, we design, implement and evaluate DFaep, a deep learning network for automatically examining, estimating, and predicting driver’s focus of attention in a real-time manner with dual low-cost dash cameras for driver-centric and car-centric views. Based on the raw stream data captured by the dash cameras during driving, we first detect the driver’s face and eye and generate augmented face images to extract facial features and enable real-time head movement tracking. We then parse the driver’s attention behaviors and gaze focus together with the road scene data captured by one front-facing dash camera faced towards the roads. Facial features, augmented face images, and gaze focus data are then inputted to our deep learning network for modeling drivers’ driving and attention behaviors. Experiments are then conducted on the large dataset, DR(eye)VE, and our own dataset under realistic driving conditions. The findings of this study indicated that the distribution of driver’s attention and focus is highly skewed. Results show that DFaep can quickly detect and predict the driver’s attention and focus, and the average accuracy of prediction is 99.38%. This will provide a basis and feasible solution with a computational learnt model for capturing and understanding driver’s attention and focus to help avoid fatal collisions and eliminate the probability of potential unsafe driving behavior in the future.

Prospective motion correction for diffusion weighted EPI of the brain using an optical markerless tracker.

PurposeTo enable motion‐robust diffusion weighted imaging of the brain using well‐established imaging techniques.MethodsAn optical markerless tracking system was used to estimate and correct for rigid body motion of the head in real time during scanning. The imaging coordinate system was updated before each excitation pulse in a single‐shot EPI sequence accelerated by GRAPPA with motion‐robust calibration. Full Fourier imaging was used to reduce effects of motion during diffusion encoding. Subjects were imaged while performing prescribed motion patterns, each repeated with prospective motion correction on and off.ResultsProspective motion correction with dynamic ghost correction enabled high quality DWI in the presence of fast and continuous motion within a 10° range. Images acquired without motion were not degraded by the prospective correction. Calculated diffusion tensors tolerated the motion well, but ADC values were slightly increased.ConclusionsProspective correction by markerless optical tracking minimizes patient interaction and appears to be well suited for EPI‐based DWI of patient groups unable to remain still including those who are not compliant with markers.

Deep Feature Attention and Aggregation for Real-Time Head Semantic Segmentation

This paper proposes a mobile CNN(Convolutional Neural Networks) architecture named FAANet(Feature Attention Aggregation Net), which focused on self-attention and feature aggregation for real time head semantic segmentation using limited computation resource. The proposed network is based on a single lightweight backbone cnn network and aggregates discriminative features through multiple different scale feature maps, and the different feature maps are fused using self-attention sub-block. Using the proposed multi-scale feature communication and propagation, FAANet substantially reduces the size of model, but still achieves big receptive field and strengthen the generalization ability, which achieves a balance between the computation cost and performance. Experiments on head segmentation dataset demonstrate the improved performance of FAANet with 10 times less FLOPs and 3times faster than the existing state-of-the-art fast semantic segmentation methods and the proposed method can achieve comparable accuracy. Specifically, when computed on one NVIDIA Titan X card, it can achieve 98.3% Mean IOU on our mixed test dataset with only 1.2 GFLOPs and 100 FPS while inferring on a 1024*1024 input image.

Medical Augmented-Reality Visualizer for Surgical Training and Education in Medicine

This paper presents a projection-based augmented-reality system (MARVIS) that supports the visualization of internal structures on the surface of a liver phantom. MARVIS is endowed with three key features: tracking of spatial relationship between the phantom and the operator’s head in real time, monoscopic projection of internal liver structures onto the phantom surface for 3D perception without additional head-mounted devices, and phantom internal electronic circuit to assess the accuracy of a syringe guidance system. An initial validation was carried out by 25 medical students (12 males and 13 females; mean age, 23.12 years; SD, 1.27 years) and 3 male surgeons (mean age, 43.66 years; SD, 7.57 years). The validation results show that the ratio of failed syringe insertions was reduced from 50% to 30% by adopting the MARVIS projection. The proposed system suitably enhances a surgeon’s spatial perception of a phantom internal structure.

Real-time Subject-specific Head and Facial Mimic Animation System using a Contactless Kinect Sensor and System of Systems Approach.

Facial palsies due to stroke, accidental and sportive injuries or sometimes without etiology, affect the professional and personal lives of involved patients. These disorders are not only a functional handicap but also a social integration impairment. The recovery of facial mimics with a normal and symmetrical facial expression allows involved patients to improve their living conditions and social identity. Current approaches lack of visual feedbacks. To monitor facial mimics and head movements in a quantitative and objective manners, a computer-aided animation system needs to be developed. Numerous systems have been proposed using single camera, stereo camera, 3-D scanner, and Kinect approaches. In particular, Kinect contactless sensor has been proven to be very suitable for 3-D facial simulation applications. However, little studies have employed the Kinect sensor for real-time head animation applications. Consequently, this study developed a real-time head and facial mimic animation system using the contactless Kinect sensor and the system of systems approach. To evaluate the accuracy of the subject-specific Kinect-based geometrical models, magnetic resonance imaging (MRI) data were used. As results, the mean distance deviation between generated Kinect-based and reconstructed MRI-based geometrical head models are approximately 1 mm for two tested subjects. The generation times are 9.7 s ± 0.3 and 0.046 s ± 0.005 by using the full facial landmarks and MPEG-4 facial landmarks respectively. Real-time head and facial mimic animations were illustrated. Particularly, the system could be executed at a very high framerate (60 fps). Further developments relate to the integration of texture information and internal structures such as a skull and muscle network to develop a full subject specific head and facial mimic animation system for facial mimic rehabilitation.

Studying Brain Function in Children Using Magnetoencephalography.

Magnetoencephalography (MEG) is a non-invasive neuroimaging technique which directly measures magnetic fields produced by the electrical activity of the human brain. MEG is quiet and less likely to induce claustrophobia compared with magnetic resonance imaging (MRI). It is therefore a promising tool for investigating brain function in young children. However, analysis of MEG data from pediatric populations is often complicated by head movement artefacts which arise as a consequence of the requirement for a spatially-fixed sensor array that is not affixed to the child's head. Minimizing head movements during MEG sessions can be particularly challenging as young children are often unable to remain still during experimental tasks. The protocol presented here aims to reduce head movement artefacts during pediatric MEG scanning. Prior to visiting the MEG laboratory, families are provided with resources that explain the MEG system and the experimental procedures in simple, accessible language. An MEG familiarization session is conducted during which children are acquainted with both the researchers and the MEG procedures. They are then trained to keep their head still whilst lying inside an MEG simulator. To help children feel at ease in the novel MEG environment, all of the procedures are explained through the narrative of a space mission. To minimize head movement due to restlessness, children are trained and assessed using fun and engaging experimental paradigms. In addition, children's residual head movement artefacts are compensated for during the data acquisition session using a real-time head movement tracking system. Implementing these child-friendly procedures is important for improving data quality, minimizing participant attrition rates in longitudinal studies, and ensuring that families have a positive research experience.

Real-Time Head Action Recognition Based on HOF and ELM

Real-Time Head Action Recognition Based on HOF and ELM

Robotic TMS mapping of motor cortex in the developing brain

BackgroundThe human motor cortex can be mapped safely and painlessly with transcranial magnetic stimulation (TMS) to explore neurophysiology in health and disease. Human error likely contributes to heterogeneity of such TMS measures. Here, we aimed to use recently pioneered robotic TMS technology to develop an efficient, reproducible protocol to characterize cortical motor maps in a pediatric population. New methodMagnetic resonance imaging was performed on 12 typically developing children and brain reconstructions were paired with the robotic TMS system. The system automatically aligned the TMS coil to target sites in 3 dimensions with near-perfect coil orientation and real-time head motion correction. Motor maps of 4 forelimb muscles were derived bilaterally by delivering single-pulse TMS at predefined, uniformly spaced trajectories across a 10 × 10 grid (7 mm spacing) customized to the participant’s MRI. ResultsProcedures were well tolerated with no adverse events. Two male, eight-year-old participants had high resting motor thresholds that precluded mapping. The mean hotspot coordinate and centre of gravity coordinate were determined in each hemisphere for four forelimb muscles bilaterally. Average mapping time was 14.25 min per hemisphere. Comparison with existing methodsTraditional manual TMS methods of motor mapping are time intensive, technically challenging, prone to human error, and arduous for use in pediatrics. This novel TMS robot approach facilitates improved efficiency, tolerability, and precision in derived, high-fidelity motor maps. ConclusionsRobotic TMS opens new avenues to explore motor map neurophysiology and its influence on developmental plasticity and therapeutic neuromodulation. Our findings provide evidence that TMS robotic motor mapping is feasible in young participants.

Postural stability predicts the likelihood of cybersickness in active HMD-based virtual reality

Cybersickness is common during virtual reality experiences with head-mounted displays (HMDs). Previously it has been shown that individual differences in postural activity can predict which people are more likely to experience visually-induced motion sickness. This study examined whether such predictions also generalise to the cybersickness experienced during active HMD-based virtual reality. Multisensory stimulation was generated by having participants continuously turn their heads from left to right while viewing the self-motion simulations. Real-time head tracking was then used to create ecological (‘compensated’) and non-ecological (‘inversely compensated’) head-and-display motion conditions. Ten (out of 20) participants reported feeling sick after being exposed to these self-motion simulations. Cybersickness did not differ significantly between the two compensation conditions. However, individual differences in spontaneous postural instability when standing quietly were found to predict the likelihood of subsequently experiencing cybersickness. These findings support recent proposals that postural measures can help diagnose who will benefit the most/least from HMD-based virtual reality.