Visual Tracking Research Articles

Dual knowledge distillation for visual tracking with teacher–student network

Dual knowledge distillation for visual tracking with teacher–student network

Visual Tracking in Amblyopia: A Continuous Psychophysical Approach.

This study aimed to explore the underlying mechanisms of the observed visuomotor deficit in amblyopia. Twenty-four amblyopic (25.8 ± 3.8years; 15 males) and 22 normal participants (25.8 ± 2.1years; 8 males) took part in the study. The participants were instructed to continuously track a randomly moving Gaussian target on a computer screen using a mouse. In experiment 1, the participants performed the tracking task at six different target sizes. In experiments 2 and 3, they were asked to track a target with the contrast adjusted to individual's threshold. The tracking performance was represented by the kernel function calculated as the cross-correlation between the target and mouse displacements. The peak, latency, and width of the kernel were extracted and compared between the two groups. In experiment 1, target size had a significant effect on the kernel peak (F(1.649, 46.170) = 200.958, P = 4.420 × 10-22). At the smallest target size, the peak in the amblyopic group was significantly lower than that in the normal group (0.089 ± 0.023 vs. 0.107 ± 0.020, t(28) = -2.390, P = 0.024) and correlated with the contrast sensitivity function (r = 0.739, P = 0.002) in the amblyopic eyes. In experiments 2 and 3, with equally visible stimuli, there were still differences in the kernel between the two groups (all Ps < 0.05). When stimulus visibility was compensated, amblyopic participants still showed significantly poorer tracking performance.

Video-tracked Anopheles arabiensis entry and exit behaviour at washed and damaged pyrethroid-treated bednets.

Insecticide-treated nets (ITNs) are the most effective method for malaria prevention in Africa. Using near-infrared video tracking in a laboratory environment, we recorded and assessed bednet entry and exit by a northern Tanzanian population of Anopheles arabiensis at a human-occupied untreated net and a PermaNet® 2.0 ITN. Both had 12 holes, each 10 cm in diameter, punctured at specific locations, and the ITN was washed 20 times to further simulate the wear and tear of ageing. Washing reduced the insecticide content of ITNs by 61%, which then showed similar rates to the untreated nets for net entry (39% entered untreated net and 41% entered ITN; p = 0.84) and exit (37% and 43%, respectively; p = 0.67). Regardless of treatment, approximately 40% of mosquitoes entered nets within 20 s of first appearing in the field of view and reached the volunteer's skin within 5 s of entering the net. Mortality rates post-exposure were significantly higher (p = 0.048) at ITNs (26.6%; 95% CI 13.4%-39.7%) than at untreated controls (6.4%; 95% CI 1.8%-14.6%). The washed and aged ITN provided little additional personal protection for the sleeper over an untreated net. Simple adjustments to materials and design that could extend the effective lifespan of ITNs are discussed.

UAV-Ground Visual Tracking: A Unified Dataset and Collaborative Learning Approach

UAV-Ground Visual Tracking: A Unified Dataset and Collaborative Learning Approach

Large-language models facilitate discovery of the molecular signatures regulating sleep and activity.

Sleep, locomotor and social activities are essential animal behaviors, but their reciprocal relationships and underlying mechanisms remain poorly understood. Here, we elicit information from a cutting-edge large-language model (LLM), generative pre-trained transformer (GPT) 3.5, which interprets 10.2-13.8% of Drosophila genes known to regulate the 3 behaviors. We develop an instrument for simultaneous video tracking of multiple moving objects, and conduct a genome-wide screen. We have identified 758 fly genes that regulate sleep and activities, including mre11 which regulates sleep only in the presence of conspecifics, and NELF-B which regulates sleep regardless of whether conspecifics are present. Based on LLM-reasoning, an educated signal web is modeled for understanding of potential relationships between its components, presenting comprehensive molecular signatures that control sleep, locomotor and social activities. This LLM-aided strategy may also be helpful for addressing other complex scientific questions.

MFACNet: A Multi-Frame Feature Aggregating and Inter-Feature Correlation Framework for Multi-Object Tracking in Satellite Videos

Efficient multi-object tracking (MOT) in satellite videos is crucial for numerous applications, ranging from surveillance to environmental monitoring. Existing methods often struggle with effectively exploring the correlation and contextual cues inherent in the consecutive features of video sequences, resulting in redundant feature inference and unreliable motion estimation for tracking. To address these challenges, we propose the MFACNet, a novel multi-frame features aggregating and inter-feature correlation framework for enhancing MOT in satellite videos with the idea of utilizing the features of consecutive frames. The MFACNet integrates multi-frame feature aggregation techniques with inter-feature correlation mechanisms to improve tracking accuracy and robustness. Specifically, our framework leverages temporal information across the features of consecutive frames to capture contextual cues and refine object representations over time. Moreover, we introduce a mechanism to explicitly model the correlations between adjacent features in video sequences, facilitating a more accurate motion estimation and trajectory associations. We evaluated the MFACNet using benchmark datasets for satellite-based video MOT tasks and demonstrated its superiority in terms of tracking accuracy and robustness over state-of-the-art performance by 2.0% in MOTA and 1.6% in IDF1. Our experimental results highlight the potential of precisely utilizing deep features from video sequences.

Physical and biophysical markers of assessment in medical training: A scoping review of the literature

Purpose Assessment in medical education has changed over time to measure the evolving skills required of current medical practice. Physical and biophysical markers of assessment attempt to use technology to gain insight into medical trainees’ knowledge, skills, and attitudes. The authors conducted a scoping review to map the literature on the use of physical and biophysical markers of assessment in medical training. Materials and methods The authors searched seven databases on 1 August 2022, for publications that utilized physical or biophysical markers in the assessment of medical trainees (medical students, residents, fellows, and synonymous terms used in other countries). Physical or biophysical markers included: heart rate and heart rate variability, visual tracking and attention, pupillometry, hand motion analysis, skin conductivity, salivary cortisol, functional magnetic resonance imaging (fMRI), and functional near-infrared spectroscopy (fNIRS). The authors mapped the relevant literature using Bloom’s taxonomy of knowledge, skills, and attitudes and extracted additional data including study design, study environment, and novice vs. expert differentiation from February to June 2023. Results Of 6,069 unique articles, 443 met inclusion criteria. The majority of studies assessed trainees using heart rate variability (n = 160, 36%) followed by visual attention (n = 143, 32%), hand motion analysis (n = 67, 15%), salivary cortisol (n = 67, 15%), fMRI (n = 29, 7%), skin conductivity (n = 26, 6%), fNIRs (n = 19, 4%), and pupillometry (n = 16, 4%). The majority of studies (n = 167, 38%) analyzed non-technical skills, followed by studies that analyzed technical skills (n = 155, 35%), knowledge (n = 114, 26%), and attitudinal skills (n = 61, 14%). 169 studies (38%) attempted to use physical or biophysical markers to differentiate between novice and expert. Conclusion This review provides a comprehensive description of the current use of physical and biophysical markers in medical education training, including the current technology and skills assessed. Additionally, while physical and biophysical markers have the potential to augment current assessment in medical education, there remains significant gaps in research surrounding reliability, validity, cost, practicality, and educational impact of implementing these markers of assessment.

Successful Management of an Acute Dystonic Reaction Induced by an Anesthetic Agent

Acute dystonic reactions (ADRs) are movement abnormalities characterized by involuntary muscle contractions that typically manifest after exposure to a triggering agent, such as a medication. The specific muscle groups affected determine the type of reaction. For instance, an oculogyric crisis primarily affects the ocular muscles, while oromandibular dystonia involves jaw opening and tongue protrusion. We present the rare case of a 68-year-old man with amyotrophic lateral sclerosis who was successfully treated for an ADR. The patient was admitted with loss of consciousness due to respiratory failure. Tracheostomy was promptly performed under sedation with multiple general anesthetic agents. Immediately after tracheostomy, the patient communicated via eye-blinking without any notable abnormalities, just as before the procedure. However, the following day, he became unresponsive to verbal cues and exhibited a decreased level of consciousness, accompanied by tongue dyskinesia, deviation of both eyes to the left, and loss of visual tracking. The patient’s vital signs remained stable. Brain imaging and an electroencephalogram revealed no abnormalities. Treatment with midazolam produced initial improvement; however, due to a significant side effect of hypotension, the treatment was switched to oral diazepam. The patient’s condition gradually improved, and the medication was eventually discontinued without further ADR episodes.

Visual control for robotic 3D printing on a moving platform

In recent years, there has been significant progress in developing specialized 3D printing techniques that cater to various demanding applications. However, the current state of this technology is challenged when it comes to complex in situ printing scenarios, which require a controlled printing platform. The lack of a stable printing platform is a fundamental limitation of its use in in situ applications. To address this issue, we present a novel platform-independent 3D fabrication process that enables printing on platforms with non-cooperative movement. The process overcomes the challenge of high-speed tracking, motion compensation, and real-time printing by developing a closed-loop visual feedback-controlled robotic printing process. The proposed process incorporates a marker-based visual detection and tracking controller setup, which is discussed in detail. The algorithm consists of two loops running asynchronously: a high-speed inner control loop and an outer measurement loop. This setup enables precise and accurate tracking of the printing platform, compensating for any disturbances during the printing process. Our experimental results demonstrate the successful printing of simple linear geometries, even with low-disturbing platform velocities. Moreover, the tracking controllers' ability to handle measurement occlusion is validated, showing the proposed process's robustness and effectiveness. Our work provides a significant step towards enabling 3D printing in complex in situ printing scenarios.

Feasibility of Computerized Visuomotor Integration System for Visual Field Defects and Spatial Neglect in Poststroke Patients.

To develop a computerized visuomotor integration system for assessment and training of visual perception impairments and evaluate its safety and feasibility in patients with a stroke. Visual field defects and spatial neglect lead to substantial poststroke impairment. Most diagnostic assessments are anchored in traditional methods, and clinical effects of rehabilitation treatments are limited. The CoTras Vision system included two evaluations and four training modules. The evaluation modules were based on the Albert's test and Star cancellation test, and training modules were based on visual tracking, central-peripheral integration, and visuomotor perception techniques. Bland-Altman plots for agreement with the traditional paper-and-pencil test were performed, and the modified Intrinsic Motivation Inventory, Patient Satisfaction Questionnaire, and Simulator Sickness Questionnaire were conducted. Ten patients with acute stroke completed the study. Bland-Altman plots revealed good agreements for Albert's test (mean difference, -0.3±4.5) and Star cancellation test (mean difference, 0.3±0.7). The mean±standard deviation scores of the modified Intrinsic Motivation Inventory, Patient Satisfaction Survey, and Simulator Sickness Questionnaire were 84.7±30.6, 40.5±7.9, and 34.0±34.5 respectively. The CoTras Vision system is feasible and safe in patients with stroke. Most patients had a high degree of motivation to use the system and did not experience severe adverse events. Further studies are needed to confirm its usefulness in stroke patients with visual field defects and hemineglect symptoms. Furthermore, a large, well-designed, randomized controlled trial will be needed to confirm the treatment effect of the CoTras Vision system.

Visual tracking with pyramidal feature fusion and transformer based model predictor

Discriminative correlation filters (DCF) have achieved much success in visual tracking. However, most of them simply rely on the features extracted by the last layer of the backbone, while ignoring the low-level rich structural information. In addition, they normally minimize the tailored objective functions to predict the target model in a direct way, which introduces inductive bias and limits the expressivity of the trackers. In view of this, a pyramidal feature fusion module is proposed in this paper to integrate the low-resolution, semantically strong features with high-resolution, semantically weak features. Then, an asymmetric Transformer structure is applied to predict the weights of the model. Finally, a feature refinement module is employed to optimize the search features. Extensive experiments on 5 mainstream datasets demonstrate the superiority of our tracker, where it has achieved better feature expression and more precise target localization.

Drone-Assisted Particulate Matter Measurement in Air Monitoring: A Patent Review

Air pollution is caused by the presence of polluting elements. Ozone (O3), carbon monoxide (CO), carbon dioxide (CO2), nitrogen dioxide (NO2), sulfur dioxide (SO2), and particulate matter (PM) are the most controlled gasses because they can be released into the atmosphere naturally or as a result of human activity, which affects air quality and causes disease and premature death in exposed people. Depending on the substance being measured, ambient air monitors have different types of air quality sensors. In recent years, there has been a growing interest in designing drones as mobile sensors for monitoring air pollution. Therefore, the objective of this paper is to provide a comprehensive patent review to gain insight into the proprietary technologies currently used in drones used to monitor outdoor air pollution. Patent searches were conducted using three different patent search engines: Google Patents, WIPO’s Patentscope, and the United States Patent and Trademark Office (USPTO). The analysis of each patent consists of extracting data that supply information regarding the type of drone, sensor, or equipment for measuring PM, the lack or presence of a cyclone separator, and the ability to process the turbulence generated by the drone’s propellers. A total of 1473 patent documents were retrieved using the search engine. However, only 13 met the inclusion criteria, including patent documents reporting drone designs for outdoor air pollution monitoring. Therefore, was found that most patents fall under class G01N (measurement; testing) according to the International Patents Classification, where the most common sensors and devices are infrared or visible light cameras, cleaning devices, and GPS tracking devices. The most common tasks performed by drones are air pollution monitoring, assessment, and control. These categories cover different aspects of the air pollution management cycle and are essential to effectively address this environmental problem.

Pupil Response in Visual Tracking Tasks: The Impacts of Task Load, Familiarity, and Gaze Position.

Pupil size is a significant biosignal for human behavior monitoring and can reveal much underlying information. This study explored the effects of task load, task familiarity, and gaze position on pupil response during learning a visual tracking task. We hypothesized that pupil size would increase with task load, up to a certain level before decreasing, decrease with task familiarity, and increase more when focusing on areas preceding the target than other areas. Fifteen participants were recruited for an arrow tracking learning task with incremental task load. Pupil size data were collected using a Tobii Pro Nano eye tracker. A 2 × 3 × 5 three-way factorial repeated measures ANOVA was conducted using R (version 4.2.1) to evaluate the main and interactive effects of key variables on adjusted pupil size. The association between individuals' cognitive load, assessed by NASA-TLX, and pupil size was further analyzed using a linear mixed-effect model. We found that task repetition resulted in a reduction in pupil size; however, this effect was found to diminish as the task load increased. The main effect of task load approached statistical significance, but different trends were observed in trial 1 and trial 2. No significant difference in pupil size was detected among the three gaze positions. The relationship between pupil size and cognitive load overall followed an inverted U curve. Our study showed how pupil size changes as a function of task load, task familiarity, and gaze scanning. This finding provides sensory evidence that could improve educational outcomes.

Adaptive and Anti-Drift Motion Constraints for Object Tracking in Satellite Videos

Object tracking in satellite videos has garnered significant attention due to its increasing importance. However, several challenging attributes, such as the presence of tiny objects, occlusions, similar objects, and background clutter interference, make it a difficult task. Many recent tracking algorithms have been developed to tackle these challenges in tracking a single interested object, but they still have some limitations in addressing them effectively. This paper introduces a novel correlation filter-based tracker, which uniquely integrates attention-enhanced bounding box regression and motion constraints for improved single-object tracking performance. Initially, we address the regression-related interference issue by implementing a spatial and channel dual-attention mechanism within the search area’s region of interest. This enhancement not only boosts the network’s perception of the target but also improves corner localization. Furthermore, recognizing the limitations in small size and low resolution of target appearance features in satellite videos, we integrate motion features into our model. A long short-term memory (LSTM) network is utilized to create a motion model that can adaptively learn and predict the target’s future trajectory based on its historical movement patterns. To further refine tracking accuracy, especially in complex environments, an anti-drift module incorporating motion constraints is introduced. This module significantly boosts the tracker’s robustness. Experimental evaluations on the SatSOT and SatVideoDT datasets demonstrate that our proposed tracker exhibits significant advantages in satellite video scenes compared to other recent trackers for common scenes or satellite scenes.

Effect of chronic stress on the degree of acrylamide toxicity in rats

Introduction. Chronic stress plays an important role in the development of a wide range of mental, somatic and behavioural disorders, and therefore is a risk factor for the health in living organisms. A negative effect on the state of the nervous system is also observed when the body is poisoned with acrylamide, which is a substance of the second class of danger and is considered toxic and carcinogenic. Exposure of the body to mental and toxic stressors leads to autonomic and neuroendocrine activation, which, in turn, manifests itself as special behavioural patterns. &#x0D; The purpose of the experiment was to assess the effect of chronic stress on the degree of acrylamide toxicity in rats.&#x0D; Materials and methods. The experiment was conducted on white outbred rats with an average weight of 200 g of both sexes (n=60), which were evenly distributed into five groups: negative control, chronic stress, acrylamide, acrylamide + treatment, acrylamide + chronic stress. Throughout the study, the animals were kept under standard conditions with 12 hours of artificial lighting during the day, a relatively constant level of humidity (30–70%) and an air temperature of 20–25 °C. Behavioural tests were conducted once a week for one calendar month using a 40×40 hole board, a multifunctional cage to assess general activity, and an elevated plus maze with an ANY-maze video tracking system. Biochemical parameters were assessed in accordance with the manufacturer’s instructions.&#x0D; Results. Analysis of behaviour on a board with holes did not reveal statistically significant results (H=8.987; p=0.061). When comparing vertical and horizontal motor activity between groups, statistically significant differences were found (p&lt;0.05). AST levels were higher in the stress groups, while cholesterol levels in the same groups were lower (p&lt;0.05).&#x0D; Limitations. Laboratory animals of only one biological species were used for the experiment. The toxicant was used only in one concentration.&#x0D; Conclusion. Chronic stress may influence on acrylamide toxicity in rats to a certain extent.

The effect of special exercises using the visual stimuli device on the speed of motor response, visual tracking, the skills of cutting and dispersing the ball, and various defensive movements for young handball players

The research is significant because it aims to improve the speed of motor response and visual tracking through targeted exercises using the visual stimuli device that the two researchers created as part of the training program for the sample. This program also supports the individual defensive movements of the players, which are exemplified by the skills of cutting and dispersing the ball and other defensive movements for young handball players. Regarding the research issue, the researchers observed a deficiency in the players' defensive performance, which manifests as slow movement and reaction to the ball's movement. This deficiency could be attributed to a lack of interest in handball player training programs as well as a weakness in visual abilities and motor response speed to the attacker's and the ball's movement. Preparing unique workouts with visual stimuli to enhance young handball players' motor reaction speed, visual tracking abilities, ball-cutting and ball-dispersing capabilities, and defensive maneuvers is one of the research's goals. The research population was purposefully selected and sampled; specifically, the goalkeepers, who numbered two, were omitted, and the youth players of the Diyala Handball Club for the 2022–2023 sports season, numbered eighteen. As a result, the sample size increased to sixteen participants, who were split into eight players, representing the experimental group, and eight players, representing the control group, at random (even and odd numbers). After conducting an exploratory experiment on two handball players, the researchers came to the conclusion that the experimental study sample's development of visual tracking and motor reaction speed was positively impacted by workouts on the visual stimulus device. The researchers suggest that in order to improve handball players' efficiency, it is imperative to engage in visual stimulus workouts that enhance skill performance and physical and visual capacities.

Multi-Object Detection using Enhanced YOLOv2 and LuNet Algorithms in Surveillance Videos

Multiple object tracking (MOT) in videos benefits multiple applications, including robot navigation, video surveillance, video analytics, and intelligent transportation systems. Although significant progress has been made since early studies, visual tracking of many objects remains challenging because of frequent occlusions in measurements, environmental noise, changeable number of objects, and appearance similarity across objects. The proposed work focused on three significant processes, feature extraction, object detection, and classification, to identify moving objects before sharing information. This work proposes a multi-object video detection method using LuNet and deep reinforcement learning. The enhanced “you only look once” version 2 (YOLOv2) initially detects numerous objects. In this work, a base network of the YOLOv2 changed by lowering the metrics and substituting it with LuNet. In the enhanced model, the LuNet network is used for feature extraction to extract the most expected characteristics from the image. Furthermore, the proposed model is compact because of the underlying network's LuNet architecture. To demonstrate the proposed technique's performance, this method compares it to numerous state-of-the-art algorithms on the MOT20 vehicle benchmark dataset. The proposed method achieves a higher % classification accuracy of 94% for moving object settings. The experiments demonstrate that the proposed method outperforms existing models in terms of performance and accuracy.

Accurate target estimation with image contents for visual tracking

Accurate target estimation with image contents for visual tracking

Search and hunt based-driving scene perception network for distance estimation and object detection

Visual tracking is generally utilized in numerous applications including surveillance and autonomous driving to provide a wide angle of vision. However, different types of alteration, including occlusion, pose, and illumination present in the video sequence act as one of the crucial troubles considered in visual tracking. Hence, this research developed a modified Deep learning (DL) approach useful for distance estimation, object detection, and pixel-wise semantic segmentation by shared convolutional architecture. The proposed object detection model employs a modified Driving Scene Perception network (DSP), which makes use of multi-task learning to enhance the efficacy of distance evaluation, semantic segmentation, and object detection. The prime significance of the research relies on the Search and Hunt optimization (SaHO) algorithm, which optimally selects the hyperparameter value of the DSP network. The experimental analysis substantiates that the proposed modified DSP network effectively estimates the distance between the moving object and the camera. The experimental analysis depicts the effectiveness of the DSP with SaHO model for both segmentation and object detection. For object detection, the DSP with SaHO model attained high rates of 0.98, 0.94, and 0.96 for accuracy, specificity, and sensitivity respectively utilizing the cityscape dataset. Additionally, the DSP with SaHO approach attained the lowest errors for segmentation and surpassed other existing techniques that revealed the efficiency of the DSP with the SaHO model.

VIDEO TRACKING BASED GRAVITY ACCELERATION ANALYSIS ON PENDULUM SWING

VIDEO TRACKING BASED GRAVITY ACCELERATION ANALYSIS ON PENDULUM SWING