Target Template Research Articles

Transformer Sub-Patch Matching for High-Performance Visual Object Tracking

Visual tracking is a core component of intelligent transportation systems, especially for unmanned driving and road surveillance. Numerous convolutional neural network (CNN) trackers have achieved unprecedented performance. However, CNN features with regular spatial context relationships experience difficulty matching the rigid target templates when dramatic deformation and occlusion occur. In this paper, we propose a novel full Transformer Sub-patch Matching network for tracking (TSMtrack), which decomposes the tracked object into sub-patches, and interlaced matches the extracted sub-patches by leveraging the attention mechanism born with the Transformer. Roots in Transformer architecture, TSMtrack consists of image patch decomposition, sub-patch matching, and position prediction. Specifically, TSMtrack converts the whole frame into sub-patches and extracts the sub-patch features independently. By sub-patch matching and FFN-like prediction, TSMtrack enables independent similarity measurement between sub-patch features in an interlaced and iterative fashion. With a full Transformer pipeline implemented, we achieve a high-quality trade-off between tracking speed performance. Experiments on nine benchmarks demonstrate the effectiveness of our Transformer sub-patch matching framework. In particular, it realizes an AO of 75.6 on GOT-10K and SR of 57.9 on WebUAV-3M with 48 FPS on GPU RTX-2060s.

Object tracking based on siamese network with 3D attention and multiple graph attention

Currently, the object tracking algorithm based on the siamese network is the most popular research direction of object tracking. However, most siamese network trackers are unable to update the template, resulting in hardly dealing with the fuzzy target well, which requires the trackers to extract the target features in the template more accurately. Moreover, most of the siamese network trackers compute the similarity between the target template and the search region in a global matching way. Thus they cannot well handle the rapid change of the target. To address the above limitations, we propose an object tracking algorithm based on 3D attention SimAM and multiple graph attention. We combine siamese networks with 3D attention to improve the feature extraction capability of the network. We frame the target template and search region to improve feature-matching accuracy by introducing multiple graph attention mechanisms. We test our model against three authoritative benchmarks, GOT10K, UAV123 and OTB100, and compare it with advanced trackers, and the results show that our trackers achieve better results.

Anti-Similar Visual Target Tracking Algorithm Based on Filter Peak Guidance and Fusion Network

Visual tracking is a key research area in computer vision, as tracking technology is increasingly being applied in daily life, it has high-research significance. Visual tracking technology usually faces various challenging interference factors, among which, a similar background is one of the factors that has a greater impact on the tracking process. Kernelized Correlation Filter (KCF) tracking algorithm can track targets quickly by using circulant matrix, and has good tracking effect, so it is widely used in the tracking field. However, when the target is interfered by similar objects, the filter template in KCF cannot effectively distinguish between the target and the interfering object. This is because the filter only uses the texture gradient feature as the description object of the target, which will make the KCF algorithm extremely sensitive to the change of the target; therefore, the filter has difficultly making a judgment in the unstable scene, cannot accurately describe the target state, and finally leads to tracking failure. Therefore, this paper fuses Color Names (CN) on the basis of the original Histogram of Oriented Gradients (HOG) feature of KCF, which can obtain a more comprehensive feature representation, and realize the application of combined features to improve the anti-interference ability of KCF in complex scenes. In addition, this paper also uses the peak response of correlation filtering as the judgment condition to determine whether the current tracking result is stable. When the filter is in an unstable tracking state, the proposed algorithm will select the value with high confidence from its multiple responses as the candidate target of the Siamese network, and the deep learning network is used as the incremental learning method of the filter. The Channel Attention is introduced into the network layer, so that the network can adaptively reason and adjust the extracted universal features, and the enhanced feature information is used as the final discriminant basis. Finally, according to the response, the target with the smallest error compared with the target template is selected from multiple candidate targets as the final tracking result. The experimental results show that the average accuracy and average success rate of the proposed algorithm are significantly improved compared with the classical tracking algorithm, especially in dealing with similar target interference.

An Overview of Paclitaxel and Molecular Imprinted Polymers Capped with Quantum Dots as an Alternative Approach for Paclitaxel Extraction and Detection

Abstract:Paclitaxel (Taxol) is a drug that belongs to the class of compounds called Taxane. It is a strong and potent chemotherapeutic drug that inhibits the growth of certain types of cancer cells; however, its abundance is very low, and various types of methodologies have been implemented to extract paclitaxel from the bark of different plants and herbs. The molecularly imprinted polymers (MIPs) could be the best alternative to purify the paclitaxel molecule. MIPs have become an attractive solution for the selective and fine-tuned determination of target templates in complex forms where other comparable and relevant structural compounds could coexist. Implementation of quantum dots in MIPs improves their extraction features due to the presence of distinct functional sites. Quantum dots can be employed to modulate the size, detectability, and state of the imprinted materials, depending on the selected application. This review aims to summarize and illustrate the modern and innovative strategies based on the aggregation of MIPs with quantum dots. Quantum dot embedded MIPs can be exploited for simultaneous extraction, preconcentration, and detection of paclitaxel obtained from various sources.

A new method for constructing discrete hyperchaotic systems with a controllable range of Lyapunov exponents and its application in information security

Since people use chaos extensively for a wide range of applications in data encryption and secure communications, a new method for designing practical high-dimensional discrete hyperchaotic systems is proposed for the first time in this paper. This method controls the range of the Lyapunov exponents in reverse by adding control variables so that the range of the values of the Lyapunov exponents is controlled within a specified interval, which is more suitable for engineering applications. Then, it is mathematically proved that the method ensures that the orbits of chaotic systems are globally finite and their Lyapunov exponents are bounded. In addition, as a practical demonstration of the selective image encryption scheme based on target template matching introduced in this paper, a 6D discrete hyperchaotic system was created, and the analysis of the simulation results verifies the applicability of the 6D hyperchaotic system designed by the method presented in this paper in the field of image encryption.

Incidental recognition reveals attentional tradeoffs shaped by categorical similarity.

Search efficiency suffers when observers look for multiple targets or a single imprecisely defined target. These conditions prevent a narrow target template, resulting in improved delayed distractor recognition. In our first experiment with hybrid visual and memory search, we investigated the interaction of target variety and target number on search efficiency. Results supported the hypothesis that numerous targets impair search efficiency much more when targets are unrelated. These efficiency impairments were linked to distractor processing, indicated by increased delayed recognition. A second experiment manipulated target-distractor similarity to determine whether prioritization of target-defining features is totally eliminated in search for eight unrelated targets. For related and unrelated targets alike, recognition declined for distractors bearing less resemblance to targets and more to each other. This suggests templates for unrelated targets successfully prioritize relevant features at some stage of attention. Avoidance of random distractors was stronger when targets were related, at the price of slower, more error-prone identification of within-category distractors. Within-category processing difficulty for related targets likely stems from categorical interference as previously demonstrated in recognition memory. Thus, target variety versus homogeneity afforded different advantages and limitations depending on target number, target-distractor, and distractor-distractor resemblance. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Robust online learning based on siamese network for ship tracking

The complex and changeable inland river scenes resulting out of frequent occlusions of ships in the available tracking methods are not accurate enough to estimate the motion state of the target ship leading to object tracking drift or even loss. In view of this, an attempt is made to propose a robust online learning ship tracking algorithm based on the Siamese network and the region proposal network. Firstly, the algorithm combines the off-line Siamese network classification score and the online classifier score for discriminative learning, and establishes an occlusion determination mechanism according to the classification the fusion score. When the target is in the occlusion state, the target template is not updated, and the global search mechanism is activated to relocate the target, thereby avoiding object tracking drift. Secondly, an efficient adaptive online update strategy, UpdateNet, is introduced to improve the template degradation in the tracking process. Finally, on comparing the state-of-the-art tracking algorithms on the inland river ship datasets, the experimental results of the proposed algorithm show strong robustness in occlusion scenarios with an accuracy and success rate of 56.8% and 57.2% respectively. Supportive source codes for this research are publicly available at https://github.com/Libra-jing/SiamOL.

Development of a CRISPR/Cas12a-recombinase polymerase amplification assay for visual and highly specific identification of the Congo Basin and West African strains of mpox virus.

Human mpox is a zoonotic disease, similar to smallpox, caused by the mpox virus, which is further subdivided into Congo Basin and West African clades with different pathogenicity. In this study, a novel diagnostic protocol utilizing clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated 12a nuclease (CRISPR/Cas12a)-mediated recombinase polymerase amplification (RPA) was developed to identify mpox in the Congo Basin and West Africa (CRISPR-RPA). Specific RPA primers targeting D14L and ATI were designed. CRISPR-RPA assay was performed using various target templates. In the designed CRISPR-RPA reaction system, the exponentially amplified RPA amplification products with a protospacer adjacent motif (PAM) site can locate the Cas12a/crRNA complex to its target regions, which successfully activates the CRISPR/Cas12a effector and achieves ultrafast trans-cleavage of a single-stranded DNA probe. The limit of detection for the CRISPR-RPA assay was 10 copies per reaction for D14L- and ATI-plasmids. No cross-reactivity was observed with non-mpox strains, confirming the high specificity of the CRISPR-RPA assay for distinguishing between the Congo Basin and West African mpox. The CRISPR-RPA assay can be completed within 45 min using real-time fluorescence readout. Moreover, the cleavage results were visualized under UV light or an imaging system, eliminating the need for a specialized apparatus. In summary, the developed CRISPR/RPA assay is a visual, rapid, sensitive, and highly specific detection technique that can be used as an attractive potential identification tool for Congo Basin and West African mpox in resource-limited laboratories.

A Computational Investigation on Rho-related GTP-binding Protein RhoB through Molecular Modeling and Molecular Dynamics Simulation Study

Background: An indispensable member of the Rho family, RhoB is an isoprenylated small GTPases that modulate the cellular cytoskeletal organization. While DNA gets damaged, it takes part in the neoplastic apoptotic mechanism. In this study, we evaluated the structure of Rho-related GTP-binding protein RhoB due to the unavailability of 3D structure in the protein data bank database. Results: The expected pI value of RhoB was 5.10 (acidic). The target–template alignment was computed using the GMQE value meanwhile 6hxu.1.A from Homo sapiens was selected as the template structure. The Swiss model was exploited to complete the model construction task. The structural compatibility and stability were revealed after a 100ns molecular dynamics simulation using GROMACA employing the OPLS-AA force field. Based on their fluctuating activity and their location between 100 and 110 and 140 and 150, PCA analysis discovered relevant residues. Conclusion: By providing an insight into the biophysical phenomenon of Rho-related GTP-binding protein RhoB inhibitors, this study will assist future investigations addressing the relationship between gene mutation and abnormalities produced by protein Rho-related GTP-binding protein RhoB in apoptotic events.

Nitrogen-Doped Hierarchical Porous Carbon Derived from Coal for High-Performance Supercapacitor.

The surface properties and the hierarchical pore structure of carbon materials are important for their actual application in supercapacitors. It is important to pursue an integrated approach that is both easy and cost-effective but also challenging. Herein, coal-based hierarchical porous carbon with nitrogen doping was prepared by a simple dual template strategy using coal as the carbon precursor. The hierarchical pores were controlled by incorporating different target templates. Thanks to high conductivity, large electrochemically active surface area (483 m2 g-1), hierarchical porousness with appropriate micro-/mesoporous channels, and high surface nitrogen content (5.34%), the resulting porous carbon exhibits a high specific capacitance in a three-electrode system using KOH electrolytes, reaching 302 F g-1 at 1 A g-1 and 230 F g-1 at 50 A g-1 with a retention rate of 76%. At 250 W kg-1, the symmetrical supercapacitor assembled at 6 M KOH shows a high energy density of 8.3 Wh kg-1, and the stability of the cycling is smooth. The energy density of the symmetric supercapacitor assembled under ionic liquids was further increased to 48.3 Wh kg-1 with a power output of 750 W kg-1 when the operating voltage was increased to 3 V. This work expands the application of coal-based carbon materials in capacitive energy storage.

Template-update-based visual tracking for automatic cell manipulation in FluidFM

Visual tracking is an essential process for performing automated cell manipulation tasks. However, Real-time accurate cell tracking in FluidFM is challenging due to probe occlusion during manipulation. This study presents a correlation filtering-based tracking method for real-time, high-precision position tracking of cells and probes. In this method, a target template update strategy is proposed to realize cell tracking under probe occlusion, which abandons the use of pixel values in the occlusion area and reduces the interference of the occlusion area in model learning. Cell tracking and automated manipulation experiments were carried out to validate the proposed method. The Experimental results show that the proposed tracking method can accurately measure the positions of occlusion cells and probe. Furthermore, combined with the precise force control of the FluidFM’s probe, the automated cell Pick & Place experiments was accomplished with high accuracy and robustness.

Good-enough attentional guidance.

Theories of attention posit that attentional guidance operates on information held in a target template within memory. The template is often thought to contain veridical target features, akin to a photograph, and to guide attention to objects that match the exact target features. However, recent evidence suggests that attentional guidance is highly flexible and often guided by non-veridical features, a subset of features, or only associated features. We integrate these findings and propose that attentional guidance maximizes search efficiency based on a 'good-enough' principle to rapidly localize candidate target objects. Candidates are then serially interrogated to make target-match decisions using more precise information. We suggest that good-enough guidance optimizes the speed-accuracy-effort trade-offs inherent in each stage of visual search.

Research on Automatic Recognition Method of Artificial Ground Target Based on Improved HED

Automatic target recognition technology is an important research direction in the field of machine vision. Artificial ground targets, such as bridges, airports and houses, are mostly composed of straight lines. The ratio of geometric primitive lines to the triangle area formed by their combination is used as the feature quantity to describe the group of lines, so as to characterize the artificial ground target. In view of the shortcomings of traditional edge detection methods, such as background suppression, non-prominent targets, missing positions, etc., this paper proposed an image edge detection method based on depth learning. By combining the traditional edge detection algorithm with the edge detection algorithm based on an improved HED network, the real-time target image edge detection was completed. An automatic target recognition method based on template matching was proposed. This method solved the problem of both homologous template matching and heterogeneous template matching, which has important theoretical value. First, the lines were combined to form the geometric primitives of the line group, and then the relationship of the lines in the group was determined by using the characteristic quantity of the line group. The best line group matching the target template was found in the image edge, and the homonymous points in the real-time image and the target template were calculated. The affine transformation matrix between the two images was obtained according to the homonymous points, and then the accurate position of the target in the real-time image was found.

Adaptive temporal feature modeling for visual tracking via cross-channel learning

Convolution Neural Networks (CNNs) based trackers achieve excellent tracking performance on tracking accuracy and speed. Feature extraction from the target template and search regions is a key part in visual tracking. Recently, existing feature subnetworks combine CNNs with channel attention for feature extraction. However, some existing feature subnetworks do not make the best of the target location dependencies, which result the target location dependency information lost in extracted target features. In this work, we design a novel feature extraction subnetwork with local temporal adaptive modules to obtain location sensitive importance maps, which effectively capture the diverse motion information and highlight the target location information. The target location dependency information is fully utilized to obtain more accurate target location information of the target template and search region in feature extraction subnetwork. The feature extraction subnetwork also fully exploits the local temporal semantics. Furthermore, we learn an interactive module in the template branch, which further captures the non-linear cross-channel interaction and channel-wise dependencies by combining every channel and its k neighbors. The template branch further utilizes cross-channel interactions for capturing the channel dependencies. The interactive module only increases a little extra computational burden. Comparing with other attention modules for visual tracking, this interactive module is lightweight. We propose a novel tracking framework, which mainly includes the designed feature extraction subnetwork and the interactive learning module. We evaluate the proposed tracker on GOT-10k, UAV123, DTB70, NFS, OTB-100, VOT2018, LaSOT and VOT-RGBT2019 benchmarks against advanced trackers, achieving leading performance with 60 FPS tracking speed.

Siamese Graph Attention Networks for robust visual object tracking

Siamese-based trackers usually convert the object tracking task into a similarity matching problem between the target template and the search region. Since fixed or manually updated templates are not robust when tracking moving objects with dramatically changing appearance, this paper proposes an improved siamese graph attention network with adaptive template update called SiamGT. By establishing spatiotemporal and context dependencies between historical images and search regions, a frame selection mechanism is added to improve the richness of information. In addition, a graph attention network with residual connections is used in the template update mechanism which enables the propagation and aggregation of information to generate robust templates. Extensive experimental results on challenging benchmarks such as UAV123, OTB100, and VOT2019 demonstrate that the proposed SiamGT has achieved state-of-the-art performance in visual object tracking.

Describing movement learning using metric learning.

Analysing movement learning can rely on human evaluation, e.g. annotating video recordings, or on computing means in applying metrics on behavioural data. However, it remains challenging to relate human perception of movement similarity to computational measures that aim at modelling such similarity. In this paper, we propose a metric learning method bridging the gap between human ratings of movement similarity in a motor learning task and computational metric evaluation on the same task. It applies metric learning on a Dynamic Time Warping algorithm to derive an optimal set of movement features that best explain human ratings. We evaluated this method on an existing movement dataset, which comprises videos of participants practising a complex gesture sequence toward a target template, as well as the collected data that describes the movements. We show that it is possible to establish a linear relationship between human ratings and our learned computational metric. This learned metric can be used to describe the most salient temporal moments implicitly used by annotators, as well as movement parameters that correlate with motor improvements in the dataset. We conclude with possibilities to generalise this method for designing computational tools dedicated to movement annotation and evaluation of skill learning.

Determination of aflatoxin B1 by biomass derived porous carbon modified electrode with molecularly imprinted polymer

AbstractAflatoxin determination is imperative among agricultural and food products as the pollution of the toxic aflatoxin could cause severe hazards towards animals and human beings. Here, we prepared Elaeagnus gum derived porous carbon material as electrode modified material. In situ electropolymerization of molecularly imprinted polymer utilizing phenol as monomer and aflatoxin B1 as target template was further processed onto the porous carbon surface to achieve aflatoxin B1 sensing ability. This sensitive MIP sensor has a detection range from 5 pM to 100 pM of aflatoxin B1, an electrochemical sensitivity of 82.4 μA log(pM)−1 cm−2 with a detection limit of 1.7 pM, and a recovery rate of real sample measurement is 98.21 %. Good selectivity, fair repeatability and stability was confirmed. This work demonstrates the promising application of Elaeagnus gum derived porous carbon modified electrode sensor for food and drug monitoring.

Target templates and the time course of distractor location learning

When searching for a shape target, colour distractors typically capture our attention. Capture is smaller when observers search for a fixed target that allows for a feature-specific target template compared to a varying shape singleton target. Capture is also reduced when observers learn to predict the likely distractor location. We investigated how the precision of the target template modulates distractor location learning in an additional singleton search task. As observers are less prone to capture with a feature-specific target, we assumed that distractor location learning is less beneficial and therefore less pronounced than with a mixed-feature target. Hierarchical Bayesian parameter estimation was used to fit fine-grained distractor location learning curves. A model-based analysis of the time course of distractor location learning revealed an effect on the asymptotic performance level: when searching for a fixed-feature target, the asymptotic distractor cost indicated smaller distractor interference than with a mixed-feature target. Although interference was reduced for distractors at the high-probability location in both tasks, asymptotic distractor suppression was less pronounced with fixed-feature compared to mixed-feature targets. We conclude that with a more precise target template less distractor location learning is required, likely because the distractor dimension is down-weighted and its salience signal reduced.

Safe and sensitive detection of ESAT6 in non-infectious clinical samples via an aptamer-based qPCR strategy for tuberculosis diagnosis

Tuberculosis (TB) is one of the leading cause of infection-related deaths worldwide. Conventional diagnostic methods for TB require infectious samples and are time-consuming. In this study, we aimed to develop an aptamer-based qPCR (Apta-qPCR) assay to diagnose TB safely and rapidly. The assay uses a single-stranded DNA aptamer with affinity for a TB biomarker, ESAT6, and molecular beacon capable of hybridizing to the aptamer. Apta-qPCR enables aptamer-mediated detection of ESAT6 via qPCR, wherein the aptamer serves as both a target recognition agent and template for quantification. This diagnostic method achieved a desirable detection limit of 2.03 nM in buffer and 2.56 nM in serum-spiked conditions with high reproducibility and reliability. In addition, the diagnostic performance was verified in a clinical application using human serum samples from normal controls and patients with active TB. The assay showed 50.0% sensitivity, 91.7% specificity, and 78.1% accuracy, similar to sputum smear microscopy, which is currently the most used method for TB diagnosis. Importantly, this assay has an advantage over smear microscopy in that it reduces the risk of infection by avoiding the use of sputum-containing infectious pathogens.

Degraded RNA from Human Anterior Cruciate Ligaments Yields Valid Gene Expression Profiles.

Correlating gene expression patterns with biomechanical properties of connective tissues provides insights into the molecular processes underlying the tissue growth and repair. Cadaveric specimens such as human knees are widely considered suitable for biomechanical studies, but their usefulness for gene expression experiments is potentially limited by the unavoidable, nuclease-mediated degradation of RNA. Here, we tested whether valid gene expression profiles can be obtained using degraded RNA from human anterior cruciate ligaments (ACLs). Human ACL RNA (N = 6) degraded in vitro by limited ribonuclease digestion resemble highly degraded RNA isolated from cadaveric tissue. PCR threshold cycle (Ct) values for 90 transcripts (84 extracellular matrix, 6 housekeeping) in degraded RNAs variably ranged higher than values obtained from their corresponding non-degraded RNAs, reflecting both the expected loss of target templates in the degraded preparations as well as differences in the extent of degradation. Relative Ct values obtained for mRNAs in degraded preparations strongly correlated with the corresponding levels in non-degraded RNA, both for each ACL as well as for the pooled results from all six ACLs. Nuclease-mediated degradation produced similar, strongly correlated losses of housekeeping and non-housekeeping gene mRNAs. RNA degraded in situ yielded comparable results, confirming that in vitro digestion effectively modeled degradation by endogenous ribonucleases in frozen and thawed ACL. We conclude that, contrary to conventional wisdom, PCR-based expression analyses can yield valid mRNA profiles even from RNA preparations that are more than 90% degraded, such as those obtained from connective tissues subjected to biomechanical studies. Furthermore, legitimate quantitative comparisons between variably degraded tissues can be made by normalizing data to appropriate housekeeping transcripts.