Target Specific Features Research Articles

Online multiple pedestrians tracking using deep temporal appearance matching association

In online multi-target tracking, modeling of appearance and geometric similarities between pedestrians visual scenes is of great importance. The higher dimension of inherent information in the appearance model compared to the geometric model is problematic in many ways. However, due to the recent success of deep-learning-based methods, handling of high-dimensional appearance information becomes feasible. Among many deep neural networks, Siamese network with triplet loss has been widely adopted as an effective appearance feature extractor. Since the Siamese network can extract the features of each input independently, one can update and maintain target-specific features. However, it is not suitable for multi-target settings that require comparison with other inputs. To address this issue, we propose a novel track appearance model based on the joint-inference network. The proposed method enables a comparison of two inputs to be used for adaptive appearance modeling and contributes to the disambiguation of target-observation matching and to the consolidation of identity consistency. Diverse experimental results support the effectiveness of our method. Our work was recognized as the 3rd-best tracker in BMTT MOTChallenge 2019, held at CVPR2019. (https://motchallenge.net/results/CVPR_2019_Tracking_Challenge/) The code is available athttps://github.com/yyc9268/Deep-TAMA.

Discriminative and Robust Online Learning for Siamese Visual Tracking

The problem of visual object tracking has traditionally been handled by variant tracking paradigms, either learning a model of the object's appearance exclusively online or matching the object with the target in an offline-trained embedding space. Despite the recent success, each method agonizes over its intrinsic constraint. The online-only approaches suffer from a lack of generalization of the model they learn thus are inferior in target regression, while the offline-only approaches (e.g., convolutional siamese trackers) lack the target-specific context information thus are not discriminative enough to handle distractors, and robust enough to deformation. Therefore, we propose an online module with an attention mechanism for offline siamese networks to extract target-specific features under L2 error. We further propose a filter update strategy adaptive to treacherous background noises for discriminative learning, and a template update strategy to handle large target deformations for robust learning. Effectiveness can be validated in the consistent improvement over three siamese baselines: SiamFC, SiamRPN++, and SiamMask. Beyond that, our model based on SiamRPN++ obtains the best results over six popular tracking benchmarks and can operate beyond real-time.

A target contrast signal theory of parallel processing in goal-directed search.

Feature Integration Theory (FIT) set out the groundwork for much of the work in visual cognition since its publication. One of the most important legacies of this theory has been the emphasis on feature-specific processing. Nowadays, visual features are thought of as a sort of currency of visual attention (e.g., features can be attended, processing of attended features is enhanced), and attended features are thought to guide attention towards likely targets in a scene. Here we propose an alternative theory - the Target Contrast Signal Theory - based on the idea that when we search for a specific target, it is not the target-specific features that guide our attention towards the target; rather, what determines behavior is the result of an active comparison between the target template in mind and every element present in the scene. This comparison occurs in parallel and is aimed at rejecting from consideration items that peripheral vision can confidently reject as being non-targets. The speed at which each item is evaluated is determined by the overall contrast between that item and the target template. We present computational simulations to demonstrate the workings of the theory as well as eye-movement data that support core predictions of the theory. The theory is discussed in the context of FIT and other important theories of visual search.

Recent advances in chronic obstructive pulmonary disease pathogenesis: from disease mechanisms to precision medicine.

Chronic obstructive pulmonary disease (COPD) is a devastating lung disease with a high personal and societal burden. Exposure to toxic particles and gases, including cigarette smoke, is the main risk factor for COPD. Together with smoking cessation, current treatment strategies of COPD aim to improve symptoms and prevent exacerbations, but there is no disease‐modifying treatment. The biggest drawback of today's COPD treatment regimen is the ‘one size fits all’ pharmacological intervention, mainly based on disease severity and symptoms and not the individual's disease pathology. To halt the worrying increase in the burden of COPD, disease management needs to be advanced with a focus on personalized treatment. The main pathological feature of COPD includes a chronic and abnormal inflammatory response within the lungs, which results in airway and alveolar changes in the lung as reflected by (small) airways disease and emphysema. Here we discuss recent developments related to the abnormal inflammatory response, ECM and age‐related changes, structural changes in the small airways and the role of sex‐related differences, which are all relevant to explain the individual differences in the disease pathology of COPD and improve disease endotyping. Furthermore, we will discuss the most recent developments of new treatment strategies using biologicals to target specific pathological features or disease endotypes of COPD. © 2020 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of Pathological Society of Great Britain and Ireland.

Multi-target regression via target specific features

Multi-target regression (MTR) has recently attracted great interest by the research community due to its capability of learning multiple related regression tasks. Existing approaches to MTR learn prediction models based on an identical set of features, which may be suboptimal since different targets may possess specific characteristics of their own. In this paper, we propose a method MTR-TSF that deals with the MTR tasks by learning target specific features (TSF). Firstly, a hierarchical clustering algorithm is applied to the output space of training data to reveal the similarities among multiple targets. Then, a classification and regression tree boosting method (CART-boosting) is used to generate a dependent similarity matrix for each target. Finally, target specific features are learned by querying the corresponding dependent similarity matrix and conducting clustering analysis on the training data. MTR-TSF method leverages the pertinent and discriminative features of each target and the dependence among multiple target to improve the overall performance of MTR. Experimental results on 18 real-world datasets demonstrate that MTR-TSF can achieve competitive performance against representative state-of-the-art MTR methods.

Usability assessment of interaction management support in LOUISE, an ECA-based user interface for elders with cognitive impairment

Embodied conversational agents (ECA) are possible enablers of assistive technologies, in particular for older adults with cognitive impairment. Yet, dedicated interaction management techniques addressing the specificities of this public are needed. OBJECTIVES: We assess whether the interaction management framework of the LOUISE (Lovely User Interface for Servicing Elders) ECA has the potential to overcome the user interface constraints linked to cognitive impairment. METHODS: LOUISE supports key target-specific features: personalization; attention management; context reminders; image and video displays; a conversation manager for task-oriented interactions; and the foundations for a domain-specific XML-based language for task-oriented assistive scenarios. LOUISE’s usability and acceptance were evaluated at the Broca geriatric hospital in Paris. with a group of 14 older adults with either mild cognitive impairment (MCI) or Alzheimer’s disease (AD) through four simple but realistic assistive scenarios: drinking, taking medicine, measuring blood pressure and choosing the lunch menu. RESULTS: Most of our participants were able to interact with the ECA, succeeded in completing the proposed tasks and enjoyed our design. CONCLUSION: The field usability evaluation of LOUISE’s interaction management framework suggests that this suite of interaction techniques can be effective in enabling interfaces for users with MCI or AD.

Scaffolding the writing development of the Argument genre in history: The case of two novice writers

We document the writing development of the Argument genre among two multilingual novice writers of academic English after they participated in a literacy intervention in a university-level history class at an English-medium university in the Middle East. We designed three writing workshops based on our previous research to target specific linguistic features of the Argument genre that were challenging for students. We draw on Systemic Functional Linguistics (SFL), which provides systematic tools for examining academic writing development, to analyze two students' essays across a semester. Our analysis shows how students who enter university with limited academic writing experiences can benefit from explicit instruction and how writing develops differently, even among students with similar skill sets. These findings support the importance of explicit disciplinary instruction (De La Paz, 2005; Monte-Sano, 2010) with an explicit focus on language (Coffin, 2006b) and contribute to the limited research on language-based writing interventions to improve student writing, particularly at the university level. Detailed qualitative analysis of these students' writing sheds light on the intricacies of writing development, which can help teachers anticipate and mitigate potential challenges.

Justice Without Borders: The Influence of Psychological Distance and Construal Level on Moral Exclusion.

The present research examines how psychological distance influences the weight given to individuating information about targets of justice judgments. Drawing on construal level theory, which links psychological distance to levels of construal, we hypothesize that increasing psychological distance from justice judgments reduces people's sensitivity to specific features of targets, thereby minimizing the extent to which applications of justice are influenced by target-specific information. Psychological proximity, by contrast, enhances the salience of targets' idiosyncratic characteristics, thereby leading to applications of justice that are more sensitive to targets' identity. Six studies, examining various justice principles, support these conclusions. Studies 1 to 3 show that psychological distancing reduces the weight of target-specific features in justice judgments. Supporting the role of construal level in driving these results, Studies 4 to 6 demonstrate parallel patterns when construal level is manipulated directly. This work offers a novel outlook on the role of construal and target characteristics in moral exclusion.

TRAP1 regulates cell cycle and apoptosis in thyroid carcinoma cells.

Tumor necrosis factor receptor-associated protein 1 (TRAP1) is a heat shock protein 90 (HSP90) molecular chaperone upregulated in several human malignancies and involved in protection from apoptosis and drug resistance, cell cycle progression, cell metabolism and quality control of specific client proteins. TRAP1 role in thyroid carcinoma (TC), still unaddressed at present, was investigated by analyzing its expression in a cohort of 86 human TCs and evaluating its involvement in cancer cell survival and proliferation in vitro Indeed, TRAP1 levels progressively increased from normal peritumoral thyroid gland, to papillary TCs (PTCs), follicular variants of PTCs (FV-PTCs) and poorly differentiated TCs (PDTCs). By contrast, anaplastic thyroid tumors exhibited a dual pattern, the majority being characterized by high TRAP1 levels, while a small subgroup completely negative. Consistently with a potential involvement of TRAP1 in thyroid carcinogenesis, TRAP1 silencing resulted in increased sensitivity to paclitaxel-induced apoptosis, inhibition of cell cycle progression and attenuation of ERK signaling. Noteworthy, the inhibition of TRAP1 ATPase activity by pharmacological agents resulted in attenuation of cell proliferation, inhibition of ERK signaling and reversion of drug resistance. These data suggest that TRAP1 inhibition may be regarded as potential strategy to target specific features of human TCs, i.e., cell proliferation and resistance to apoptosis.

Diversity of LFPs Activated in Different Target Regions by a Common CA3 Input.

Identifying the pathways contributing to local field potential (LFP) events and oscillations is essential to determine whether synchronous interregional patterns indicate functional connectivity. Here, we studied experimentally and numerically how different target structures receiving input from a common population shape their LFPs. We focused on the bilateral CA3 that sends gamma-paced excitatory packages to the bilateral CA1, the lateral septum, and itself (recurrent input). The CA3-specific contribution was isolated from multisite LFPs in target regions using spatial discrimination techniques. We found strong modulation of LFPs by target-specific features, including the morphology and population arrangement of cells, the timing of CA3 inputs, volume conduction from nearby targets, and co-activated inhibition. Jointly they greatly affect the LFP amplitude, profile, and frequency characteristics. For instance, ipsilateral (Schaffer) LFPs occluded contralateral ones, and septal LFPs arise mostly from remote sources while local contribution from CA3 input was minor. In the CA3 itself, gamma waves have dual origin from local networks: in-phase excitatory and nearly antiphase inhibitory. Also, waves may have different duration and varying phase in different targets. These results indicate that to explore the cellular basis of LFPs and the functional connectivity between structures, besides identifying the origin population/s, target modifiers should be considered.

Unique PTSD Clusters Predict Intention to Seek Mental Health Care and Subsequent Utilization in US Veterans with PTSD Symptoms

Many veterans return from deployment with posttraumatic stress disorder (PTSD), but most attend only a limited number of mental health care visits. Although global PTSD relates to seeking mental health care, it is unclear whether specific features of PTSD inform the low rates of mental health care utilization. This study examined PTSD cluster severities of avoidance, reexperiencing, dysphoria, and hyperarousal as predictors of intention to seek mental health care and prospective treatment utilization. US veterans with at least subthreshold PTSD (N = 189) completed a PTSD symptom measure and indicated whether they intended to seek mental health care. Prospective Department of Veterans Affairs mental health care utilization was extracted from the medical record. At the bivariate level, each cluster was positively associated with a positive intention to seek mental health care and prospective treatment utilization. In multivariate models, however, dysphoria severity (OR = 1.16, 95% CI [1.06, 1.26]) was uniquely and positively correlated with intention to seek mental health care, whereas higher avoidance severity (IRR = 0.86, 95% CI [0.76, 0.98]) predicted lower treatment utilization, and higher reexperiencing severity (IRR = 1.07, 95% CI [1.01, 1.14]) predicted greater treatment utilization. It is critical to tailor interventions to target specific features of PTSD and to meet patients where they are.

Targeting pediatric cancer stem cells with oncolytic virotherapy

Cancer stem cells (CSC), also termed “cancer initiating cells” or “cancer progenitor cells”, which have the ability to self-renew, proliferate, and maintain the neoplastic clone, have recently been discovered in a wide variety of pediatric tumors. These CSC are thought to be responsible for tumorigenesis, tumor maintenance, aggressiveness and recurrence due to inherent resistance to current treatment modalities such as chemotherapy and radiation. Oncolytic virotherapy offers a novel, targeted approach for eradicating pediatric CSC by utilizing mechanisms of cell killing that differ from conventional therapies. Moreover, oncolytic viruses have the ability to target specific features of CSC such as cell surface proteins, transcription factors, and the CSC microenvironment. Through genetic engineering, a wide variety of foreign genes may be expressed by oncolytic viruses to augment the oncolytic effect. We review the current data regarding the ability of several types of oncolytic viruses (herpes simplex virus-1 (HSV-1), adenovirus, reovirus, Seneca Valley virus, vaccinia virus, Newcastle disease virus, myxoma virus, vesicular stomatitis virus) to target and kill both CSC and tumor cells in pediatric tumors. We highlight advantages and limitations of each virus and potential ways next-generation engineered viruses may target resilient CSC.

Relationship between working memory capacity and contingent involuntary orienting

Individual differences in working memory capacity are related to the ability to control attention; where less working memory capacity is associated with less attentional control. A well-known demonstration of attentional control is contingent involuntary orienting (Folk, Remington, & Johnston, 1992), which is the finding that visuospatial attention is captured by a salient peripheral cue, but only if the cue is featurally similar to the target of search and not if the cue is featurally dissimilar. This control of attention and ability to resist shifting attention towards an irrelevant visual cue is believed to be moderated by attentional settings for target-specific features. This study establishes that working memory capacity is related to contingent involuntary orienting of attention. Lower working memory capacity was associated with larger stimulus-driven cueing effects at short cue-to-target onset asynchronies, than at longer asynchronies, suggesting working memory capacity was associated with the ability to control covert orienting of attention at early processing stages.

On the generality of the displaywide contingent orienting hypothesis: Can a visual onset capture attention without top-down control settings for displaywide onset?

The contingent orienting hypothesis (Folk, Remington, & Johnston, 1992) states that attentional capture is contingent on top-down control settings induced by task demands. Past studies supporting this hypothesis have identified three kinds of top-down control settings: for target-specific features, for the strategy to search for a singleton, and for visual features in the target display as a whole. Previously, we have found stimulus-driven capture by onset that was not contingent on the first two kinds of settings (Yeh & Liao, 2008). The current study aims to test the third kind: the displaywide contingent orienting hypothesis (Gibson & Kelsey, 1998). Specifically, we ask whether an onset stimulus can still capture attention in the spatial cueing paradigm when attentional control settings for the displaywide onset of the target are excluded by making all letters in the target display emerge from placeholders. Results show that a preceding uninformative onset cue still captured attention to its location in a stimulus-driven fashion, whereas a color cue captured attention only when it was contingent on the setting for displaywide color. These results raise doubts as to the generality of the displaywide contingent orienting hypothesis and help delineate the boundary conditions on this hypothesis.

Target-Specific Short-Term Dynamics Are Important for the Function of Synapses in an Oscillatory Neural Network

Short-term dynamics such as facilitation and depression are present in most synapses and are often target-specific even for synapses from the same type of neuron. We examine the dynamics and possible functions of two synapses from the same presynaptic neuron in the rhythmically active pyloric network of the spiny lobster. Using simultaneous recordings, we show that the synapses from the lateral pyloric (LP) neuron to the pyloric dilator (PD; a member of the pyloric pacemaker ensemble) and the pyloric constrictor (PY) neurons both show short-term depression. However, the postsynaptic potentials produced by the LP-to-PD synapse are larger in amplitude, depress less, and recover faster than those produced by the LP-to-PY synapse. The main function of the LP-to-PD synapse is to slow down the pyloric rhythm. However, in some cases, it slows down the rhythm only when it is fast and has no effect or to speeds up when it is slow. In contrast, the LP-to-PY synapse functions to delay the activity of the PY neuron; this delay increases as the cycle period becomes longer. Using a computational model, we show that the short-term dynamics of synaptic depression observed for each of these synapses are tailored to their individual functions and that replacing the dynamics of either synapse with the other would disrupt these functions. Together, the experimental and modeling results suggest that the target-specific features of short-term synaptic depression are functionally important for synapses efferent from the same presynaptic neuron.

Implementation of algorithms for extracting tonal components in underwater noise

Underwater target classification and tracking problems utilize the noise signals emanating from the targets as well as the target dynamics features. Target classification is carried out by extracting certain classification clues about the target such as its emission frequencies, and other target specific features from the self noise or active transmissions. The self noise generated by the targets being nonstationary in nature are to be analyzed with short term averaged data segments for extracting the tonal components using various spectral estimation techniques. Though classical spectral analysis techniques are computationally efficient, it suffers from several inherent limitations such as frequency resolution, performance degradation due to implicit windowing of the data, etc. In an attempt to improve the spectral resolution, several modern spectral estimation techniques, which utilize the procedures of indirect Fourier analysis by fitting the measured short data segments to an assumed model, have been evolved. This paper presents the development of procedures for the estimation of power spectral densities using modern techniques based on parameter estimation such as auto regressive, moving average, auto regressive moving average, maximum likelihood, etc. The performance of the estimator has been validated by computing the emission frequencies of a 50-foot vessel.

Development of an underwater target classifier using target specific features

In Sonar, the detection and estimation functions are performed by signal processors, which involve the computation of various statistics, for enhancing the overall performance of the system. This also takes into account all the undesirable propagation effects caused by the underwater channel. Underwater targets can be classified by using certain target specific features such as target strength, target dynamics, and the signatures of the noise generated by the targets. Rough identification of the targets is carried out with target strength values at known aspects while for precise identification, classification clues from target dynamics and target signatures are generated. Databases for the engine noise spectra of various underwater targets, propeller noises, machinery noises and cavitation noises, speed-noise characteristics, etc., have been developed. The signal energy estimated within a finite-time interval is compared with the earlier detection/estimation decisions, which are stored in the target data record and the relevant target data are updated. The algorithm for identification of target from the most matching signature patterns in the database will generate the classification clues, which will help in target identification. Salient highlights of an underwater target classifier using the above-discussed target specific features are presented in this paper.

Assessing camouflage methods using textural features

Developments in the area of signature suppression make it progressively more difficult to recognize targets. The emphasis has been on the reduction of distinct features, like hot spots in the infrared band. Thus, to obtain a low false alarm rate, threat sensors have to utilize more information, i.e., spatial and spectral properties. The purpose of our work is to develop a general tool for camouflage assessment. The approach proposed applies texture descriptors to quantify the similarity between different parts of an image. In addition, other descriptors are used to distinguish man-made object characteristics. The selection of an appropriate set of features is discussed. The assumption is that an area containing observable targets has different statistics than other areas. Statistical properties along with detected target specific features have to be combined with methods used in data fusion. An experiment with a dataset of 44 reference images has been carried out, using a recently developed computer program called Terrtex. High correlation with perception experiments is achieved using only one or two texture features. The importance of a careful selection of background area size is finally discussed. (C) 2001 Society of Photo-optical instrumentation Engineers.

Tunable and efficient diode-pumped Tm3+-doped fluoride fiber laser for hydrocarbon gas sensing

The operation of an efficient diode laser pumped continuous wave thulium-doped fiber laser emitting at wavelength λ = 2.31 μm is reported. The operation of the fiber laser is optimized with a view to producing a small and efficient laser source for optical absorption-based gas sensing. A 2-mW output power combined with a high slope efficiency make this fiber laser a useful source for sensing spectroscopic-based hydrocarbon gases that absorb around λ = 2.3 μm. A lower limit of detection of 1000 ppm. meters for methane is routinely demonstrated using an unoptimized Fabry-Perot cavity laser system. The tunability of the fiber laser has also been investigated in order to target specific hydrocarbon absorption features. Initial results with a diffraction grating tuned system suggest detectivities of < 150 ppm.m may be obtained.

The interaction of hydrogenic ions with metal and semiconductor surfaces

We present K-Auger spectra arising from collisions of hydrogenic C, N, O, F and Ne ions with p-conducting Si(100), Ni(110) and polycrystalline W surfaces for different beam energies and observation angles. These spectra exhibit sharp peaks which, using Hartree-Fock atomic structure calculations, can be assigned to distinct projectile states prior to KLL decay. Moreover, we find that the measured spectra show target-specific features which can be explained by different free electron densities in metals and silicon.