Tracking Information Research Articles

Locuszoomr: an R package for visualising publication-ready regional gene locus plots

Abstract Summary Locuszoomr is an R package for visualising and creating publication-ready regional gene locus plots similar to those produced by the original web interface ‘LocusZoom’, but running locally in R. Genetic or genomic data with gene annotation tracks are plotted via R base graphics or ‘ggplot2’, allowing flexibility and easy customisation. Modular plotting functions enable layering of multiple GWAS plots such as for PheWAS, comparing GWAS with eQTL signals, and aligning multiple locus plots on the same page. Interactive plots can be visualised using ‘plotly’ enabling quick identification of data points for labelling. The ‘LDlink’ API (https://ldlink.nih.gov/) can be programmatically queried to obtain linkage disequilibrium (LD) data from the 1000 Genomes Project which is overlaid on plots. The package allows users to query Ensembl databases from Bioconductor or AnnotationHub for genomic information, so it can be used to show gene track information for any species available in Ensembl. Helper functions enable recombination rate data from UCSC to be overlaid on plots. The package has a detailed vignette and includes embedded example GWAS data to allow users to try out different features. Locuszoomr is also fast: identifying 50 peaks in a GWAS of around 8 million SNPs takes around 0.6 seconds. Exporting 50 locus plots with 1Mbp window with recombination rate takes around 30 seconds on Intel 9th gen CPU or 14 seconds on Apple ARM M3. Availability and implementation The R package locuszoomr is open-source and available from CRAN: https://cran.r-project.org/package=locuszoomr and GitHub: https://github.com/myles-lewis/locuszoomr

Multi-Object Tracking with Predictive Information Fusion and Adaptive Measurement Noise

Multi-object tracking (MOT) aims to detect objects in video sequences and associate them across frames. Currently, the mainstream research direction regarding MOT is the tracking-by-detection (TBD) framework. Tracking results are highly sensitive to detection outputs, and challenges from object occlusion and complex motion present significant obstacles in the field of MOT. To reduce dependence on detection outputs, we propose a method that integrates predictive information to improve Non-Maximum Suppression (NMS). By applying secondary modulation to the suppression scores and dynamically adjusting the suppression threshold using tracking information, our method better retains candidate boxes for occluded objects. Furthermore, to track occluding and overlapping objects more effectively, we introduce an adaptive measurement noise method that adjusts the measurement noise to mitigate the impact of object occlusion or overlap on tracking accuracy. Additionally, we enhance the affinity matrix in the association algorithm by incorporating height information, thereby improving the stability of complex moving objects. Our method outperforms the baseline model ByteTrack on the DanceTrack dataset, increasing Higher Order Tracking Accuracy (HOTA), Multi-Object Tracking Accuracy (MOTA), and the ID F1 Score (IDF1) by 10.2%, 3.0%, and 4.8%, respectively.

Intelligent optoelectrowetting digital microfluidic system for real-time selective parallel manipulation of biological droplet arrays.

Optoelectrowetting technology generates virtual electrodes to manipulate droplets by projecting optical patterns onto the photoconductive layer. This method avoids the complex design of the physical circuitry of dielectricwetting chips, compensating for the inability to reconstruct the electrode. However, the current technology relies on operators to manually position the droplets, draw optical patterns, and preset the droplet movement paths. It lacks real-time feedback on droplet information and the ability for independent droplet control, which can lead to droplet miscontrol and contamination. This paper presents a combination of optoelectrowetting with deep learning algorithms, integrating software and a photoelectric detection platform, and develops an optoelectrowetting intelligent control system. First, a target detection algorithm identifies droplet characteristics in real-time and automatically generate virtual electrodes to control movement. Simultaneously, a tracking algorithm outputs trajectories and ID information for efficient droplet arrays tracking. The results show that the system can automatically control the movement and fusion of multiple droplets in parallel and realize the automatic arrangement and storage of disordered droplet arrays without any additional electrodes and sensing devices. Additionally, through the automated control of the system, the cell suspension can be precisely cultured in the specified medium according to experimental requirements, and the growth trend is consistent with that observed in the well plate, significantly enhancing the experiment's flexibility and accuracy. In this paper, we propose an intelligent method applicable to the automated manipulation of discrete droplets. This method would play a crucial role in advancing the applications of digital microfluidic technology in biomedicine and other fields.

Augmented reality navigation in orthognathic surgery: Comparative analysis and a paradigm shift.

The emergence of augmented reality (AR) in surgical procedures could significantly enhance accuracy and outcomes, particularly in the complex field of orthognathic surgery. This study compares the effectiveness and accuracy of traditional drilling guides with two AR-based navigation techniques: one utilizing ArUco markers and the other employing small-workspace infrared tracking cameras for a drilling task. Additionally, an alternative AR visualization paradigm for surgical navigation is proposed that eliminates the potential inaccuracies of image detection using headset cameras. Through a series of controlled experiments designed to assess the accuracy of hole placements in surgical scenarios, the performance of each method was evaluated both quantitatively and qualitatively. The findings reveal that the small-workspace infrared tracking camera system is on par with the accuracy of conventional drilling guides, hinting at a promising future where such guides could become obsolete. This technology demonstrates a substantial advantage by circumventing the common issues encountered with traditional tracking systems and surpassing the accuracy of ArUco marker-based navigation. These results underline the potential of this system for enabling more minimally invasive interventions, a crucial step towards enhancing surgical accuracy and, ultimately, patient outcomes. The study resulted in three relevant contributions: first, a new paradigm for AR visualization in the operating room, relying only on exact tracking information to navigate the surgeon is proposed. Second, the comparative analysis marks a critical step forward in the evolution of surgical navigation, paving the way for integrating more sophisticated AR solutions in orthognathic surgery and beyond. Finally, the system with a robotic arm is integrated and the inaccuracies present in a typical human-controlled system areevaluated.

An inter-comparison of tropical cyclone datasets for the Australian region

Abstract Tropical cyclone (TC) best track datasets have temporal inhomogeneity, mostly associated with changes in monitoring practices and technological improvements. Temporal inconsistencies are often mitigated by using TC data from more homogeneous periods. For example, TC records since 1980 are preferred for frequency and track analysis, while records for intensity analysis have become more consistent since ~2000. However, such measures reduce the sample size for trend analysis, potentially leading to conflicting conclusions due to natural climate-variability. Inter-agency best track data can also vary, due to differences in the way best track information – such as centre fix locations and associated intensity estimates – are defined and assessed. When comparing global datasets and regional datasets, additional inconsistencies can be introduced where TCs form or track just outside the official area of responsibility for each agency. We highlight discrepancies in Australian TC best track data from various sources by comparing it to a more rigorously scrutinized dataset compiled by the Australian Bureau of Meteorology. This dataset is found to have highly accurate TC records for the Australian region. We also highlight the implications of data differences on TC-related trend analysis, aiming to increase awareness of dataset inconsistencies while guiding credible climate-change detection and attribution messages.

Cellular Phone Number Tracking System Based on ArcGIS and Google Maps

Over 4 billion individuals are projected to own a mobile device, with smartphone use continuing to rise around the world. Associated with this trend is the collection and usage of location information provided by these devices. The procedure for tracking and evaluating a mobile phone number is crucial for crime and terrorist attacks preventing concerns, especially in countries with deteriorated security conditions. The locational tracking information can enable real-time operational awareness and can save people's lives. In this aspect, the Geographic Information System (GIS) system is frequently used to provide intensive spatial information along the tracking process. This research aims to examine, follow up on, and trace a certain person's phone number inside a certain area using a combination of ArcGIS and Google Maps application software they are programs used to track and determine locations through the use of various types of coordinates, including latitude and longitude. In the ArcGIS program, there are several methods and algorithms used in various fields, including (locating people, determining tourist places, determining water and sewage networks, forecasting in the event of storms or floods, and many other areas), an example of these methods in the program is (Kriging method, Idw method, Spline method) and many other methods, each method is used for a specific purpose and field. In this research, the kriging method was used, which is a design algorithm that is used in many analysis and tracking processes and gives clear and accurate results. The process of determining locations in the maps used in the search depends on taking the longitude and latitude coordinates and downloading them to those maps. The maps that have been worked on in the research provide a set of basic concepts adopted in the process of tracking a particular person’s phone number, such as (the person’s location, the numbers connected to him, and the duration of a call to numbers in his contact list). The tracking system used is tested by creating a virtual contact list for a specific person’s phone number and making use of all the data in the contact list and applying it in the ArcGIS program, knowing that the duration of the contact list used is 9 months (from January to October 2021). The obtained results are analyzed and sorted in different ways to provide a wide model of the locational communication and movement activity of the tracked person. They showed that the system is efficient and can be applied to enhance the security in Iraq for tracking suspected people in crimes like terrorism, kidnapping, drugs, extortion … etc.

Advancing Public Health Interventions: A Novel Surveillance System for Hazardous Consumer Products.

Lead poisoning remains a significant public health concern with preventable exposure from different sources, including certain traditional consumer products. To address the concern of product-related lead exposures, the New York City Department of Health and Mental Hygiene has developed the Non-Paint Sample Database (NPSD). NPSD systematically tracks information about hazardous consumer products collected during New York City case investigations and store surveys. NPSD has enabled identification of new products-related lead sources and populations at risk of lead exposure and has guided risk communication, regulatory activities, and policy initiatives. Since 2007, approximately 8000 consumer products have been tracked by NPSD. NPSD demonstrates the potential for consumer product surveillance to enable source identification, monitor lead exposure trends, and facilitate effective educational and enforcement activities, which advances lead exposure prevention strategies.

An open and interoperable Knowledge Management framework for Climate Change Vulnerability Assessment

Climate change vulnerability assessment (CCVA) is crucial to understand how different regions, communities, and ecosystems are susceptible to the impacts of climate change. Through the identification of vulnerabilities, policymakers can prioritize adaptation efforts, allocate resources effectively, and develop risk mitigation strategies. Social and environmental scientists can design solutions for the development of climate-resilient regions. Nowadays, various approaches have been made available for CCVA, leading to increased knowledge along with the identification of challenges and pitfalls. Participatory approaches are required for the proper selection of vulnerability indicators and the interpretation of the assessment results, accompanied by the development of quality data assessment processes to reduce inconsistencies and uncertainties in the provided data. Openness and interoperability of data and processes have to be promoted to enable the extensibility, reproducibility and comparability of the produced results. In the current work, we develop a framework to enable open and interoperable CCVA, taking advantage of a knowledge management infrastructure and specifically a Knowledge Graph that tracks information for various indicators around the Sustainable Development Goals. We follow an indicator-based vulnerability assessment approach for various temporal and spatial scales, while considering hazard-independent and hazard-dependent analysis. Indicative analysis results are provided in national, regional and local scale.

Fault detection of key parts of wind turbine based on BP neural network combination prediction model

A BP neural network incorporated regression forecast technique based upon fragment swarm optimization (PSO) is proposed to design the state of crucial components of wind turbine so regarding realize mistake identification and detection. Firstly, specification recognition is carried out on the collection and tracking information of the system, and parameters connected to fault detection are extracted. Then, the residual optimization issue is made use of to establish the forecast model of nonlinear state evaluation and semantic network combination, and the gearbox temperature level or generator bearing are input as criteria right into the semantic network combination model and single model specifically, and the precision of the design is mirrored by the examination index. Lastly, BP design and PSO-BP combined forecast model are developed respectively by using the actual operation data of wind ranch SCADA, and the mistake state is evaluated according to whether the anticipated residual exceeds the set threshold, so regarding keep an eye on the temperature level of wind turbine transmission and generator bearing. By contrasting the data videotaped prior to and after the failing and making the information prediction analysis, the speculative results show that the forecast model established in this paper is viable for the device element fault detection.

Unsupervised learning analysis on the proteomes of Zika virus.

The Zika virus (ZIKV), which is transmitted by mosquito vectors to nonhuman primates and humans, causes devastating outbreaks in the poorest tropical regions of the world. Molecular epidemiology, supported by clustering phylogenetic gold standard studies using sequence data, has provided valuable information for tracking and controlling the spread of ZIKV. Unsupervised learning (UL), a form of machine learning algorithm, can be applied on the datasets without the need of known information for training. In this work, unsupervised Random Forest (URF), followed by the application of dimensional reduction algorithms such as principal component analysis (PCA), Uniform Manifold Approximation and Projection (UMAP), t-distributed stochastic neighbor embedding (t-SNE), and autoencoders were used to uncover hidden patterns from polymorphic amino acid sites extracted on the proteome ZIKV multi-alignments, without the need of an underlying evolutionary model. The four UL algorithms revealed specific host and geographical clustering patterns for ZIKV. Among the four dimensionality reduction (DR) algorithms, the performance was better for UMAP. The four algorithms allowed the identification of imported viruses for specific geographical clusters. The UL dimension coordinates showed a significant correlation with phylogenetic tree branch lengths and significant phylogenetic dependence in Abouheif's Cmean and Pagel's Lambda tests (p value < 0.01) that showed comparable performance with the phylogenetic method. This analytical strategy was generalizable to an external large dengue type 2 dataset. These UL algorithms could be practical evolutionary analytical techniques to track the dispersal of viral pathogens.

How Much Hatchery-Reared Brown Trout Move in a Large, Deep Subalpine Lake? An Acoustic Telemetry Study

Fish movement into large, deep lakes has been rarely investigated due to the complexity and extent of such ecosystems. Among the different monitoring methods available, acoustic telemetry enables the study of the spatial ecology and behavior of aquatic organisms in lentic environments. In this study, the movement of 69 hatchery-reared adult brown trout (size 43–61 cm) marked with acoustic transmitters was monitored in the large and deep subalpine Lake Lugano (Switzerland and Italy). Trout were tracked for six consecutive months by seven acoustic receivers (March–August 2022), positioned in a non-overlapping array. Trout movement was reconstructed using R packages specific for acoustic telemetry (actel and RSP), which also allowed us to translate tracking information into utilization distribution (UD) areas for each fish. The effects of different environmental variables (rainfall, water discharge of the two main tributaries of Lake Lugano, atmospheric pressure, cloud coverage, and moon phases) on trout movement were tested, but none of these variables seemed to significantly correlate with fish movement. After release, most of the tagged fish exhibited reiterative movements during the initial month, with some maintaining this behavior throughout the entire study period. This spatial behavior can be particularly evident in hatchery-reared fish due to their aggressive and bold attitude. The association of these behavioral traits, shaped by domestication, could expose hatchery-reared fish to high risks and post-release mortality in the wild. Indeed, within a few months after the release, most of the tagged fish were no longer detected by the acoustic receivers. In addition, 26% of the total tagged fish were caught by recreational or professional fishermen.

Development of quantitative PET/MR imaging for measurements of hepatic portal vein input function: a phantom study

BackgroundAccurate pharmacokinetic modelling in PET necessitates measurements of an input function, which ideally is acquired non-invasively from image data. For hepatic pharmacokinetic modelling two input functions need to be considered, to account for the blood supply from the hepatic artery and portal vein. Image-derived measurements at the portal vein are challenging due to its small size and image artifacts caused by respiratory motion. In this work we seek to demonstrate, using phantom experiments, how a dedicated PET/MR protocol can tackle these challenges and potentially provide input function measurements of the portal vein in a clinical setup.MethodsA custom 3D printed PET/MR phantom was constructed to mimic the liver and portal vein. PET/MR acquisitions were made with emulated respiratory motion. The PET/MR imaging protocol consisted of high-resolution anatomical MR imaging of the portal vein, followed by a PET acquisition in parallel to a dedicated motion-tracking MR sequence. Motion tracking and deformation information were extracted from PET data and subsequently used in PET reconstruction to produce dynamic series of motion-free PET images. Anatomical MR images were used post PET reconstruction for partial volume correction of the input function measurements.ResultsReconstruction of dynamic PET data with motion-compensation provided nearly motion-free series of PET frame data, suitable for image derived input function measurements of the portal vein. After partial volume correction, the individual input function measurements were within a 16.1% error range from the true activity in the portal vein compartment at the time of PET acquisition.ConclusionThe proposed protocol demonstrates clinically feasible PET/MR imaging of the liver for pharmacokinetic studies with accurate quantification of the portal vein input function, including correction for respiratory motion and partial volume effects.

End-to-end multiple object tracking in high-resolution optical sensors of drones with transformer models

Drone aerial imaging has become increasingly important across numerous fields as drone optical sensor technology continues to advance. One critical challenge in this domain is achieving both accurate and efficient multi-object tracking. Traditional deep learning methods often separate object identification from tracking, leading to increased complexity and potential performance degradation. Conventional approaches rely heavily on manual feature engineering and intricate algorithms, which can further limit efficiency. To overcome these limitations, we propose a novel Transformer-based end-to-end multi-object tracking framework. This innovative method leverages self-attention mechanisms to capture complex inter-object relationships, seamlessly integrating object detection and tracking into a unified process. By utilizing end-to-end training, our approach simplifies the tracking pipeline, leading to significant performance improvements. A key innovation in our system is the introduction of a trajectory detection label matching technique. This technique assigns labels based on a comprehensive assessment of object appearance, spatial characteristics, and Gaussian features, ensuring more precise and logical label assignments. Additionally, we incorporate cross-frame self-attention mechanisms to extract long-term object properties, providing robust information for stable and consistent tracking. We further enhance tracking performance through a newly developed self-characteristics module, which extracts semantic features from trajectory information across both current and previous frames. This module ensures that the long-term interaction modules maintain semantic consistency, allowing for more accurate and continuous tracking over time. The refined data and stored trajectories are then used as input for subsequent frame processing, creating a feedback loop that sustains tracking accuracy. Extensive experiments conducted on the VisDrone and UAVDT datasets demonstrate the superior performance of our approach in drone-based multi-object tracking.

Considerations on time resolution of neutron irradited single pixel 3D structures at fuences up to 1017 neq/cm2 using 120 GeV SPS pion beams

The proven radiation hardness of silicon 3D devices up to fluences of 1 × 1017 neq/cm2 makes them an excellent choice for next generation trackers, providing <10 μm position resolution at a high multiplicity environment. The anticipated pile-up increase at HL-LHC conditions and beyond, requires the addition of <50 ps per hit timing information to successfully resolve displaced and primary vertices. In this study, the timing performance, uniformity and efficiency of neutron and proton irradiated single pixel 3D devices is discussed. Fluences up to 1 × 1017 neq/cm2 in three different geometrical implementations are evaluated using 120 GeV SPS pion beams. A MIMOSA-26 type telescope is used to provide detailed tracking information with a ∼5 μm position resolution. Productions with single- and double-sided processes, yielding active thicknesses of 130 and 230 μm respectively, are examined with varied pixel sizes from 55 × 55 μm2 to 25 × 100 μm2 and a comparative study of field uniformity is presented with respect to electrode geometry. The question of electronics bandwidth is extensively addressed with respect to achievable time resolution, efficiency and collected charge, forming a 3D phase space to which an appropriate operating point can be selected depending on the application requirements.

Semantic Segmentation Model-Based Boundary Line Recognition Method for Wheat Harvesting

The wheat harvesting boundary line is vital reference information for the path tracking of an autonomously driving combine harvester. However, unfavorable factors, such as a complex light environment, tree shade, weeds, and wheat stubble color interference in the field, make it challenging to identify the wheat harvest boundary line accurately and quickly. Therefore, this paper proposes a harvest boundary line recognition model for wheat harvesting based on the MV3_DeepLabV3+ network framework, which can quickly and accurately complete the identification in complex environments. The model uses the lightweight MobileNetV3_Large as the backbone network and the LeakyReLU activation function to avoid the neural death problem. Depth-separable convolution is introduced into Atrous Spatial Pyramid Pooling (ASPP) to reduce the complexity of network parameters. The cubic B-spline curve-fitting method extracts the wheat harvesting boundary line. A prototype harvester for wheat harvesting boundary recognition was built, and field tests were conducted. The test results show that the wheat harvest boundary line recognition model proposed in this paper achieves a segmentation accuracy of 98.04% for unharvested wheat regions in complex environments, with an IoU of 95.02%. When the combine harvester travels at 0~1.5 m/s, the normal speed for operation, the average processing time and pixel error for a single image are 0.15 s and 7.3 pixels, respectively. This method could achieve high recognition accuracy and fast recognition speed. This paper provides a practical reference for the autonomous harvesting operation of a combine harvester.

Group activity recognition using unreliable tracked pose

Group activity recognition in video is a complex task due to the need for a model to recognise the actions of all individuals in the video and their complex interactions. Recent studies propose that optimal performance is achieved by individually tracking each person and subsequently inputting the sequence of poses or cropped images/optical flow into a model. This helps the model to recognise what actions each person is performing before they are merged to arrive at the group action class. However, all previous models rely highly on high-quality tracking and have only been evaluated using ground truth tracking information. In practice, achieving highly reliable tracking information for all individuals in a group activity video is almost impossible. We introduce an innovative deep learning-based group activity recognition approach called Rendered Pose-based Group Activity Recognition System (RePGARS), designed to tolerate unreliable tracking and pose information. Experimental results confirm that RePGARS outperforms all existing group activity recognition algorithms tested, which do not use ground truth detection and tracking information.

Microbubble tracking based on partial smoothing-based adaptive generalized labelled Multi-Bernoulli filter for super-resolution imaging

Super-resolution ultrasound (SRUS) can image the vasculature at microscopic resolution according to microbubble (MB) localization, with velocity vector maps obtained based on MB tracking information. High MB concentrations can reduce the acquisition time of SRUS imaging, however adjacent and intersecting vessels are difficult to distinguish, thus decreasing resolution. Low acquisition frame rates affect the precision of flow velocity estimation. This study proposes a partial smoothing-based adaptive generalized labeled multi-Bernoulli filter (SAGLMB) to precisely track the MB motion at different flow velocities. SAGLMB employs a generalized labelled multi-Bernoulli filter (GLMB) for MB trajectory allocation to separate adjacent and intersecting vessels. Furthermore, the nonlinear motion of MB was predicted by an unscented Kalman filter, and a cardinalized probability hypothesis density filter was applied to suppress clutter interference. Finally, the trajectories were smoothed by unscented Rauch-Tung-Striebel to improve the resolution of the SRUS image. The simulation results demonstrate that SAGLMB outperforms the conventional bipartite graph-based tracking at high MB concentrations, achieving at least an 8.55 % improvement in the correctly paired precision, with 3 times increase in the structural similarity index measure. Moreover, SAGLMB can obtain more precise flow velocity estimations with a 4 times improvement than the conventional method. The SRUS results of rabbit kidney show that the proposed method significantly improves resolution of adjacent and intersecting vessels at higher MB concentrations and maintains this performance as the acquisition frame rate decreases. Furthermore, the rat brain microvascular network was reconstructed with 9.21 μm (λ/11.1) resolution. Therefore, SAGLMB can achieve robust SRUS imaging at high concentrations and low acquisition frame rates.

Personnel Monitoring in Shipboard Surveillance Using Improved Multi-Object Detection and Tracking Algorithm.

Detecting and tracking personnel onboard is an important measure to prevent ships from being invaded by outsiders and ensure ship security. Ships are characterized by more cabins, numerous equipment, and dense personnel, so there are problems such as unpredictable personnel trajectories, frequent occlusions, and many small targets, which lead to the poor performance of existing multi-target-tracking algorithms on shipboard surveillance videos. This study conducts research in the context of onboard surveillance and proposes a multi-object detection and tracking algorithm for anti-intrusion on ships. First, this study designs the BR-YOLO network to provide high-quality object-detection results for the tracking algorithm. The shallow layers of its backbone network use the BiFormer module to capture dependencies between distant objects and reduce information loss. Second, the improved C2f module is used in the deep layer of BR-YOLO to introduce the RepGhost structure to achieve model lightweighting through reparameterization. Then, the Part OSNet network is proposed, which uses different pooling branches to focus on multi-scale features, including part-level features, thereby obtaining strong Re-ID feature representations and providing richer appearance information for personnel tracking. Finally, by integrating the appearance information for association matching, the tracking trajectory is generated in Tracking-By-Detection mode and validated on the self-constructed shipboard surveillance dataset. The experimental results show that the algorithm in this paper is effective in shipboard surveillance. Compared with the present mainstream algorithms, the MOTA, HOTP, and IDF1 are enhanced by about 10 percentage points, the MOTP is enhanced by about 7 percentage points, and IDs are also significantly reduced, which is of great practical significance for the prevention of intrusion by ship personnel.

SiamMAF: A multipath and feature-enhanced thermal infrared tracker

Thermal infrared (TIR) images are visually blurred and low in information content. Some TIR trackers focus on enhancing the semantic information of TIR features, neglecting the equally important detailed information for TIR tracking. After target localization, detailed information can assist the tracker in generating accurate prediction boxes. In addition, simple element-wise addition is not a way to fully utilize and fuse multiple response maps. To address these issues, this study proposes a multipath and feature-enhanced Siamese tracker (SiamMAF) for TIR tracking. We design a feature-enhanced module (FEM) based on complementarity, which can highlight the key semantic information of the target and preserve the detailed information of objects. Furthermore, we introduce a response fusion module (RFM) that can adaptively fuse multiple response maps. Extensive experimental results on two challenging benchmarks show that SiamMAF outperforms many existing state-of-the-art TIR trackers and runs at a steady 31FPS.

Enhanced keypoint information and pose-weighted re-ID features for multi-person pose estimation and tracking

Enhanced keypoint information and pose-weighted re-ID features for multi-person pose estimation and tracking