Ambiguous Representation Research Articles

Simultaneous multi-person tracking and activity recognition based on cohesive cluster search

We present a bootstrapping framework to simultaneously improve multi-person tracking and activity recognition at individual, interaction and social group activity levels. The inference consists of identifying trajectories of all pedestrian actors, individual activities, pairwise interactions, and collective activities, given the observed pedestrian detections. Our method uses a graphical model to represent and solve the joint tracking and recognition problems via three stages: (i) activity-aware tracking, (ii) joint interaction recognition and occlusion recovery, and (iii) collective activity recognition.This full-stack problem induces great complexity in learning the representations for the sub-problems at each stage, and the complexity increases as with more stages in the system. Our solution is to make use of symbolic cues for inference at higher stages, inspired by the observations of cohesive clusters at different stages. This also avoids learning more ambiguous representations in the higher stages.High-order correlations among the visible and occluded individuals, pairwise interactions, groups, and activities are then solved using the cohesive cluster search within a Bayesian framework. Experiments on several benchmarks show the advantages of our approach over the existing methods.

The aperture problem for binaural hearing can explain ambiguous lateralization of complex binaural stimuli

Although we perceive a spatially contiguous, integrated world, sensory systems sample information discretely via small spatial and narrow frequency receptive fields (RFs) in vision and audition, respectively. Undersampling of stimuli by narrow RFs produces ambiguous representations (e.g., the classic “aperture problem” that visual motion cannot be uniquely determined by one RF). Here, we posit a similar problem for encoding low-frequency (<1500 Hz) interaural time differences (ITDs). Current models of ITD processing propose large ensembles of ITD-sensitive neurons spanning wide ranges of preferred ITD and characteristic frequency (CF). To account for anomalous lateralization of narrowband but robust lateralization of broadband stimuli, such models require two paradoxical assumptions that have recently come into question; (1) the existence of neurons preferring larger than physiologically possible ITDs, and (2) population biases overemphasizing small ITDs. Here, we consider an alternative account based on a narrow and weighted ITDxCF aperture through which the stimuli are perceived. ITDs are bounded by ±π interaural phase difference and CFs by dominance weighting, where ∼600–750 Hz contributes most to lateralization. A more detailed consideration of this “binaural aperture” sheds light on how lateralization data emerge from limited ensembles of neurons constrained to physiologically relevant ranges of ITD and frequency.

IndelPost: harmonizing ambiguities in simple and complex indel alignments.

Small insertions and deletions (indels) in nucleotide sequence may be represented differently between mapping algorithms and variant callers, or in the flanking sequence context. Representational ambiguity is especially profound for complex indels, complicating comparisons between multiple mappings and call sets. Complex indels may additionally suffer from incomplete allele representation, potentially leading to critical misannotation of variant effect. We present indelPost, a Python library that harmonizes these ambiguities for simple and complex indels via realignment and read-based phasing. We demonstrate that indelPost enables accurate analysis of ambiguous data and can derive the correct complex indel alleles from the simple indel predictions provided by standard small variant detectors, with improved performance over a specialized tool for complex indel analysis. indelPost is freely available at: https://github.com/stjude/indelPost. Supplementary data are available at Bioinformatics online.

Elizabeth I’s Royal Progresses: A Study of Formal Orations and Poetic Recitations

The present study examines the rhetoric of inconsistency in the representation of Queen Elizabeth I through a reading of formal orations and poetic recitations written during her royal progresses. These literary resources, which were deliberately designed to promote the Elizabethan monarchy, offer illuminating examples of Elizabeth’s struggle to cultivate a distinctive royal identity. I would suggest that the tactical rhetorical practice of creating paradoxical images was an essential constituent of Elizabeth’s statecraft to cement her authority and reinforce her legitimacy. Indeed, the deployment of a discourse of contradiction that shaped Elizabeth’s progresses was a necessary and practical approach to overcome the vulnerability of an unmarried female monarch. The analysis of contradictory imagery is a valuable contribution to comprehend the complexity of Elisabeth’s representation and her strategies of exercising power in a patriarchal society. The research shows that Elizabeth employed the medium of creating ambiguous images as a rhetorical tactic to overcome gender bias against the female monarchy, and her courtiers utilized the same approach to advance their own agendas. It explores two ambiguous representations: masculine/ feminine portrayal and virgin/ maternal depiction.

25-LB: Are You a Good Steward of Data?

The Problem: There is an urgent need for innovative solutions for new and improved treatments and therapies. Often clinical research data is “silo-ed” within an organization, preventing the broader research community from leveraging aggregated information from individuals and studies globally. This isolation of data limits the discoveries possible within a disease preventing the discovery of new scientific links between disease areas. Inconsistent and ambiguous data representation leads to a lack of transparency, traceability and reproducibility, in turn fostering inefficiencies and inaccuracies in research that loses time and increases costs. Simply put Are we being good stewards of our study data? Data Standards and Data Sharing as part of the solution: Describing data consistently using data standards can bring clarity to data that amplifies its value. The T1D Data Standards Project builds on existing CDISC diabetes standards by describing concepts relevant to T1D in the areas of: • Pediatrics and Devices (Released Sep 2020) • Exercise and Nutrition (Due Q2-2021) • Screening, Staging and Monitoring for Pre-Clinical Type 1 Diabetes (Due Q3-2021). The team of CDISC standards experts, T1D experts from CDISC member organizations and T1D subject matter experts from academia and industry followed a consensus-driven process to ensure the final standards are fit for use within the research community, reinforced by the open community review. How data standards help overcome some of the challenges: Publicizing the availability and promoting adoption of consistent standards brings clarity to data, reduces data collection, organization and analysis time, and facilitates interoperability that promotes data re-use and operational efficiencies. When researchers speak the same harmonized language, talking and understanding each other is much easier leading to the sharing of common, understandable data within the research community, thereby promoting scientific and therapeutic advances general confidence in the data. Disclosure R. Baker: None. J. M. Owen: None. Funding The Leona M. and Harry B. Helmsley Charitable Trust (G-2018PG-T1D069)

Neural Decoding Reveals Concurrent Phonemic and Subphonemic Representations of Speech Across Tasks.

Robust and efficient speech perception relies on the interpretation of acoustically variable phoneme realizations, yet prior neuroimaging studies are inconclusive regarding the degree to which subphonemic detail is maintained over time as categorical representations arise. It is also unknown whether this depends on the demands of the listening task. We addressed these questions by using neural decoding to quantify the (dis)similarity of brain response patterns evoked during two different tasks. We recorded magnetoencephalography (MEG) as adult participants heard isolated, randomized tokens from a /ba/-/da/ speech continuum. In the passive task, their attention was diverted. In the active task, they categorized each token as ba or da. We found that linear classifiers successfully decoded ba vs. da perception from the MEG data. Data from the left hemisphere were sufficient to decode the percept early in the trial, while the right hemisphere was necessary but not sufficient for decoding at later time points. We also decoded stimulus representations and found that they were maintained longer in the active task than in the passive task; however, these representations did not pattern more like discrete phonemes when an active categorical response was required. Instead, in both tasks, early phonemic patterns gave way to a representation of stimulus ambiguity that coincided in time with reliable percept decoding. Our results suggest that the categorization process does not require the loss of subphonemic detail, and that the neural representation of isolated speech sounds includes concurrent phonemic and subphonemic information.

Reusability and composability in process description maps: RAS-RAF-MEK-ERK signalling.

Detailed maps of the molecular basis of the disease are powerful tools for interpreting data and building predictive models. Modularity and composability are considered necessary network features for large-scale collaborative efforts to build comprehensive molecular descriptions of disease mechanisms. An effective way to create and manage large systems is to compose multiple subsystems. Composable network components could effectively harness the contributions of many individuals and enable teams to seamlessly assemble many individual components into comprehensive maps. We examine manually built versions of the RAS–RAF–MEK–ERK cascade from the Atlas of Cancer Signalling Network, PANTHER and Reactome databases and review them in terms of their reusability and composability for assembling new disease models. We identify design principles for managing complex systems that could make it easier for investigators to share and reuse network components. We demonstrate the main challenges including incompatible levels of detail and ambiguous representation of complexes and highlight the need to address these challenges.

Swine producer willingness to pay for Tier 1 disease risk mitigation under multifaceted ambiguity

AbstractUnderstanding producers' risk perceptions of foreign animal diseases and their willingness to participate in prevention efforts is crucial for effective policymaking and communication strategies, yet there is limited empirical evidence to support initiatives. Using results from a 2017 survey of U.S. swine producers, we test whether producers' subjective risk perceptions affect their willingness to pay for interventions that could reduce potential foreign animal disease losses. Then, controlling for subjective risk perceptions, we determine the relative and total impacts of different representations of ambiguity on willingness to pay. Our results indicate that producers have a mean willingness to pay of $0.581 per pig per year. We find no difference in willingness to pay under alternative representations of ambiguity characterized by probability of occurrence, damage, and damage duration. Rather, a producer with a higher subjective belief in the probability of a foreign animal disease outbreak in the U.S. swine industry, a higher level of formal education, operating as a contractor or integrator, and owning a truck‐wash has a higher willingness to pay. Also, the number of market hogs sold annually is an important determinant of willingness to pay for foreign animal disease risk mitigation. [EconLit citations: D81, Q11, Q18]

Metastable shock waves in the region of their ambiguous representation

The behavior of perturbed shocks belonging to Hugoniot segments with ambiguous representation of the shock-wave discontinuity is considered. The segments are adjacent to the region of the shock wave instability L > 1 + 2M from the side of greater and lower pressure. Here, L is the D’yakov parameter and M is the post-shock Mach number. The model equation of state is used. For the constructed Hugoniot curve, the ambiguity sections are overlapped by the shock wave neutral stability regions. The three-dimensional simulations with varying post-shock parameters and perturbation intensity are performed. It is shown that shock waves considered are metastable: for any intensity of the shock wave, there are intervals of the perturbation threshold values inside of which the shock is neutrally stable or unstable. Neutral stability is characterized by the weak damping of secondary waves occurring as result of the perturbation.

Operational, gauge-free quantum tomography

As increasingly impressive quantum information processors are realized in laboratories around the world, robust and reliable characterization of these devices is now more urgent than ever. These diagnostics can take many forms, but one of the most popular categories istomography, where an underlying parameterized model is proposed for a device and inferred by experiments. Here, we introduce and implement efficient operational tomography, which uses experimental observables as these model parameters. This addresses a problem of ambiguity in representation that arises in current tomographic approaches (thegauge problem). Solving the gauge problem enables us to efficiently implement operational tomography in a Bayesian framework computationally, and hence gives us a natural way to include prior information and discuss uncertainty in fit parameters. We demonstrate this new tomography in a variety of different experimentally-relevant scenarios, including standard process tomography, Ramsey interferometry, randomized benchmarking, and gate set tomography.

Strategies of ideological polarisation in the online news media: A social actor analysis of Megawati Soekarnoputri

This article examines strategies of ideological polarisation in the discourse of the Indonesian online news media site, Kompas.com. Applying Van Leeuwen’s model of social actor analysis and van Dijk’s concept of the ideological square, the study focuses on the representation of Megawati Soekarnoputri, leader of the Indonesian Democratic Party of Struggle (PDI-P) as an icon of ideological contestation during the 2014 presidential election. Situated in the era of digital platform convergence, the analysis uncovers a pattern of strategically ambiguous representations of Megawati and her apparently transgressive actions and interactions. This practice entices readers to ‘read between the lines’ and activate their ideological repertoire to determine in-group and out-group members. It also enables Kompas.com to pursue commercial objectives and navigate journalistic constraints by obscuring explicitly ideological content. The implications are discussed in terms of the impact of online news media discourse upon democratic political engagement, and women’s political participation.

Chinese Clinical Named Entity Recognition in Electronic Medical Records: Development of a Lattice Long Short-Term Memory Model With Contextualized Character Representations.

BackgroundClinical named entity recognition (CNER), whose goal is to automatically identify clinical entities in electronic medical records (EMRs), is an important research direction of clinical text data mining and information extraction. The promotion of CNER can provide support for clinical decision making and medical knowledge base construction, which could then improve overall medical quality. Compared with English CNER, and due to the complexity of Chinese word segmentation and grammar, Chinese CNER was implemented later and is more challenging.ObjectiveWith the development of distributed representation and deep learning, a series of models have been applied in Chinese CNER. Different from the English version, Chinese CNER is mainly divided into character-based and word-based methods that cannot make comprehensive use of EMR information and cannot solve the problem of ambiguity in word representation.MethodsIn this paper, we propose a lattice long short-term memory (LSTM) model combined with a variant contextualized character representation and a conditional random field (CRF) layer for Chinese CNER: the Embeddings from Language Models (ELMo)-lattice-LSTM-CRF model. The lattice LSTM model can effectively utilize the information from characters and words in Chinese EMRs; in addition, the variant ELMo model uses Chinese characters as input instead of the character-encoding layer of the ELMo model, so as to learn domain-specific contextualized character embeddings.ResultsWe evaluated our method using two Chinese CNER datasets from the China Conference on Knowledge Graph and Semantic Computing (CCKS): the CCKS-2017 CNER dataset and the CCKS-2019 CNER dataset. We obtained F1 scores of 90.13% and 85.02% on the test sets of these two datasets, respectively.ConclusionsOur results show that our proposed method is effective in Chinese CNER. In addition, the results of our experiments show that variant contextualized character representations can significantly improve the performance of the model.

Uncertainty analysis of hierarchical granular structures for multi-granulation typical hesitant fuzzy approximation space

Hierarchical structures and uncertainty measures are two main aspects in granular computing, approximate reasoning and cognitive process. Typical hesitant fuzzy sets, as a prime extension of fuzzy sets, are more flexible to reflect the hesitance and ambiguity in knowledge representation and decision making. In this paper, we mainly investigate the hierarchical structures and uncertainty measures in typical hesitant fuzzy backgrounds. Firstly, we propose the parameterized scalar cardinalities of typical hesitant fuzzy elements, typical hesitant fuzzy sets and typical hesitant fuzzy relations based on a more reasonable partial orders with a disjunctive semantic meaning, respectively, where the parameters represent the decision makers' risk preferences. Secondly, we present four ordered relations for typical hesitant fuzzy space and four uncertainty measures to characterize the ambiguity in typical hesitant fuzzy approximation space and discuss their relationships. Thirdly, the hierarchical structures of a multi-granulation typical hesitant fuzzy space are analyzed by various multi-granulation typical hesitant fuzzy knowledge bases. In addition, we construct the framework of multi-granulation typical hesitant fuzzy rough sets in terms of optimistic and pessimistic attitudes. Finally, we study the uncertainty measures for the multi-granulation typical hesitant fuzzy approximation space based on the maximal and minimal knowledge bases, respectively.

In this very uncertain space

As a journalist for The Globe and Mail for 10 years, El Akkad has reported on war and conflict from around the world, including the war in Afghanistan, the military trials at Guantanamo Bay, the Arab Spring revolution in Egypt, and the Black Lives Matter movement in Ferguson, Missouri. In 2018, El Akkad's debut novel, American War, was shortlisted for a number of prominent literary prizes, garnering public attention as a finalist on the CBC Canada Reads competition. In this conversation, we talk to El Akkad about his novel, journalism, literary influences, migrations, and political visions of the future. One recurring theme in the conversation is the relationship between violence and the production of uncertainty—the unpredictability movement and refuge for the displaced; the ambiguity and risks of racial representation; the secrecy of detention and redaction; and the uncertainties of the future in times of change and crisis.

Ensemble sparse supervised model for bearing fault diagnosis in smart manufacturing

Abstract Machinery fault diagnosis is of great significance to improve the reliability of smart manufacturing. Deep learning based fault diagnosis methods have achieved great success. However, the features extracted by different models may vary resulting in ambiguous representation of the data, and even wasted time with manually selecting the optimal hyperparameters. To solve the problems, this paper proposes a new framework named Ensemble Sparse Supervised Model (ESSM), in which a typical deep learning model is treated as two phases of feature learning and model learning. In the feature learning phase, the original data is represented to be a feature matrix as non-redundant as possible by applying sparse filtering. Then, the feature matrix is fed into the model learning phase. Regularization, dropout and rectified linear unit (ReLU) are used in the model's neurons and layers to build a sparse deep neural network. Finally, the output of the sparse deep neural network provides feedback to the first phase to obtain better sparse features. In the proposed method, hyperparameters need to be pre-specified and a python library of talos is employed to finish the process automatically. The proposed method is verified using the bearing data provided by Case Western Reserve University. The result demonstrates that the proposed method can capture the effective pattern of data with the help of sparse constraints and simultaneously provide convenience for the operators with assuring performance.

Interaction of metastable shock waves with perturbations

The shock wave metastable behavior in equilibrium media is detected in numerical experiments. This phenomenon is directly related to the ambiguous representation of a shock wave discontinuity when the shock belongs to the S-shaped fragment of the pressure-velocity Hugoniot. After interaction with weak perturbations the metastable shock wave recovers its original form and parameters (like stable shock). If the amplitude of the perturbations exceeds some critical value, the shock wave instability is developed. The instability is accompanied by formation of the transverse secondary waves switching local post-shock parameters between two Hugoniot segments separated from each other by the absolute instability region. The problem of the metastable shock wave interaction with local entropy perturbation passing through the shock front is studied in the framework of three-dimensional problem formulation. The shock wave transition to the limiting self-oscillating mode is described. Some features of this mode are completely determined by the shape of the Hugoniot and isentropes and does not depend on the shock wave structure. Namely, the shape of the shock front corrugations and flow structure evolution are analyzed. These features may be served as marker of the corresponding thermodynamic anomalies.

Ambiguity Coding Allows Accurate Inference of Evolutionary Parameters from Alignments in an Aggregated State-Space.

How can we best learn the history of a protein’s evolution? Ideally, a model of sequence evolution should capture both the process that generates genetic variation and the functional constraints determining which changes are fixed. However, in practical terms the most suitable approach may simply be the one that combines the convenience of easily available input data with the ability to return useful parameter estimates. For example, we might be interested in a measure of the strength of selection (typically obtained using a codon model) or an ancestral structure (obtained using structural modeling based on inferred amino acid sequence and side chain configuration).But what if data in the relevant state-space are not readily available? We show that it is possible to obtain accurate estimates of the outputs of interest using an established method for handling missing data. Encoding observed characters in an alignment as ambiguous representations of characters in a larger state-space allows the application of models with the desired features to data that lack the resolution that is normally required. This strategy is viable because the evolutionary path taken through the observed space contains information about states that were likely visited in the “unseen” state-space. To illustrate this, we consider two examples with amino acid sequences as input. We show that n}{}omega, a parameter describing the relative strength of selection on nonsynonymous and synonymous changes, can be estimated in an unbiased manner using an adapted version of a standard 61-state codon model. Using simulated and empirical data, we find that ancestral amino acid side chain configuration can be inferred by applying a 55-state empirical model to 20-state amino acid data. Where feasible, combining inputs from both ambiguity-coded and fully resolved data improves accuracy. Adding structural information to as few as 12.5% of the sequences in an amino acid alignment results in remarkable ancestral reconstruction performance compared to a benchmark that considers the full rotamer state information. These examples show that our methods permit the recovery of evolutionary information from sequences where it has previously been inaccessible. [Ancestral reconstruction; natural selection; protein structure; state-spaces; substitution models.]

Visual Research Methods: Qualifying and Quantifying the Visual

Abstract The role of visual research methods in ethnographic research has been significant, particularly in place-making and representing visual culture and environments in ways that are not easily substituted by text. Digital media has extended into mundane, everyday existences and routines through most noticeably the modern smartphone, social media and digital artefacts that have created new forms of ethnographic enquiry. Ethnographers have engaged in this relatively new possibility of exploring how social media and new technologies transform the way we view social realities through the digital experience. The paper discusses the possible role of visual research methods in multimethod research and the theoretical underpinning of interpreting visual data. In the process of interpreting and analysing visual data, there is a need to acknowledge the possible ambiguity and polysemic quality of visual representation. It presents selectively the use of visual methods in an ethnographic exploration of early childhood settings through the use of internet-based visual data, researcher and participant-generated visual materials and media, together with visual-elicited (e.g. drawings, still images, video clips) information data through several examples. This approach in ‘visualizing’ the curriculum also unveils some aspects of the visual culture or the ‘hidden curriculum’ in the learning environment.

Difficulties with design: The challenges of teaching design in K-5 programming

Teachers in England are required to ensure that learners from the age of five are taught about algorithms and program design. Yet, there is evidence that despite teachers reporting that design is important, they are not converting this into classroom practice. This paper describes a survey study, in which we explored teachers’ difficulties in using design. We surveyed 207 teachers asking them free-text questions on their use of design in teaching programming and their views of pupils’ responses to using design. In the survey, we also investigated teachers’ understanding of the term algorithm, an essential concept which may be a contributing factor in their difficulties with design. We provide underpinning data on the difficulties of using design that teachers of pupils aged from 5 to 11 years old (Grades K to 5) have in teaching programming. Difficulties with design identified include pupil resistance, a lack of time, a lack of pupil and teacher expertise, conflicting pedagogical choices and a general confusion over what an algorithm is. There were statistically significant differences in selection of the term ‘algorithm’ to describe programming artefacts whether a teacher was a specialist or a generalist, what training they had received on programming or design, the age group taught and programming language used. Teachers were more likely to call a complex code snippet an ‘algorithm’ than a simpler one and more likely to select the term to describe code snippets than a design artefact. We make suggestions of how to alleviate the problems including that teachers are introduced to the idea of ambiguous representations of algorithms and a process which refines the representation from ambiguous to unambiguous as the design progresses.

Managing structural variability in agent-based models with feature coherence graphs

Exploratory modelling requires not only parametric but also structural and representational variability. The ability to systematically vary and experiment with multiple model features is paramount for problems that involve uncertainty and representational ambiguity. To facilitate such variation, a coherence-driven strategy is introduced for the management and evaluation of feature selections. The feature coherence specification and its associated metaprogramming system open new avenues for programmable abductive model building, as well as model introspection while automating the selection of model variants in the context of an evolving analysis. As learning takes place through simulation experiments, an increasingly accurate feature coherence model emerges to serve as an explanatory model that reveals which features cohere and are conducive to generating targeted systemic behaviours.