Emotion Recognition Robust to Indoor Environmental Distortions and Non-targeted Emotions Using Out-of-distribution Detection

The rapid development of machine learning on acoustic signal processing has resulted in many solutions for detecting emotions from speech. Early works were developed for clean and acted speech and for a fixed set of emotions. Importantly, the datasets and solutions assumed that a person only exhibited one of these emotions. More recent work has continually been adding realism to emotion detection by considering issues such as reverberation, de-amplification, and background noise, but often considering one dataset at a time, and also assuming all emotions are accounted for in the model. We significantly improve realistic considerations for emotion detection by (i) more comprehensively assessing different situations by combining the five common publicly available datasets as one and enhancing the new dataset with data augmentation that considers reverberation and de-amplification, (ii) incorporating 11 typical home noises into the acoustics, and (iii) considering that in real situations a person may be exhibiting many emotions that are not currently of interest and they should not have to fit into a pre-fixed category nor be improperly labeled. Our novel solution combines CNN with out-of-data distribution detection. Our solution increases the situations where emotions can be effectively detected and outperforms a state-of-the-art baseline.

Uncovering variability in children's concepts and conceptual change

Capturing the structure and development of human conceptual knowledge is a challenging but fundamental task in Cognitive Science. The most prominent approach to uncovering these concepts is Multidimensional scaling (MDS), which has provided insight into the structure of human perception and conceptual knowledge. However, MDS usually requires participants to produce large numbers of similarity judgments, leading to prohibitively long experiments for most developmental research. Furthermore, MDS provides a single psychological space, tailored to a fixed set of stimuli. In contrast, we present a method that learns psychological spaces flexibly and generalizes to novel stimuli. In addition, our approach uses a simple, developmentally appropriate task, which allows for short and engaging developmental studies. We evaluate the feasibility of our approach on simulated data and find that it can uncover the true structure even when the data consists of aggregations of diverse categorizers. We then apply the method to data from the World Color Survey and find that it can discover language-specific color organization. Finally, we use the method in a novel developmental experiment and find age-dependent differences in conceptual spaces for fruit categories. These results suggest that our method is robust and widely applicable in developmental tasks with children as young as four years old.

Context-bounded verification of thread pools

Thread pooling is a common programming idiom in which a fixed set of worker threads are maintained to execute tasks concurrently. The workers repeatedly pick tasks and execute them to completion. Each task is sequential, with possibly recursive code, and tasks communicate over shared memory. Executing a task can lead to more new tasks being spawned. We consider the safety verification problem for thread-pooled programs. We parameterize the problem with two parameters: the size of the thread pool as well as the number of context switches for each task. The size of the thread pool determines the number of workers running concurrently. The number of context switches determines how many times a worker can be swapped out while executing a single task---like many verification problems for multithreaded recursive programs, the context bounding is important for decidability. We show that the safety verification problem for thread-pooled, context-bounded, Boolean programs is EXPSPACE-complete, even if the size of the thread pool and the context bound are given in binary. Our main result, the EXPSPACE upper bound, is derived using a sequence of new succinct encoding techniques of independent language-theoretic interest. In particular, we show a polynomial-time construction of downward closures of languages accepted by succinct pushdown automata as doubly succinct nondeterministic finite automata. While there are explicit doubly exponential lower bounds on the size of nondeterministic finite automata accepting the downward closure, our result shows these automata can be compressed. We show that thread pooling significantly reduces computational power: in contrast, if only the context bound is provided in binary, but there is no thread pooling, the safety verification problem becomes 3EXPSPACE-complete. Given the high complexity lower bounds of related problems involving binary parameters, the relatively low complexity of safety verification with thread-pooling comes as a surprise.

An Empirical Validation of the Regression Point Displacement Design Using Within-Study Comparison Logic: Emerging Possibilities and Cautions

Background The regression point displacement (RPD) design is a quasi-experiment (QE) that aims to control many threats to internal validity. Though it has existed for several decades, RPD has only recently begun to answer applied research questions in lieu of stronger QEs. Objectives Our primary objective was to implement within-study comparison (WSC) logic to create RPD replicates and to determine conditions under which RPD might provide estimates comparable to those found in validating experiments. Research Design We utilize three randomized controlled trials (two cluster-level, one individual-level), artificially decomposing or creating cluster structures, to create multiple RPDs. We compare results in each RPD treatment group to a fixed set of control groups to gauge the congruence of these repeated RPD realizations with results found in these three RCTs. Results RPD’s performance was uneven. Using multiple criteria, we found that RPDs successfully predicted the direction of the RCT’s intervention effect but inconsistently fell within the .10 SD threshold. A scant 13% of RPD results were statistically significant at either the .05 or .01 alpha-level. RPD results were within the 95% confidence interval of RCTs around half the time, and false negative rates were substantially higher than false positive rates. Conclusions RPD consistently underestimates treatment effects in validating RCTs. We analyze reasons for this insensitivity and offer practical suggestions to improve the chances RPD will correctly identify favorable results. We note that the synthetic, “decomposition of cluster RCTs,” WSC design represents a prototype for evaluating other QEs.

Multi-Channel Local Binary Pattern Guided Convolutional Neural Network for Breast Cancer Classification

Background: The advancement in convolutional neural network (CNN) has reduced the burden of experts using the computer-aided diagnosis of human breast cancer. However, most CNN networks use spatial features only. The inherent texture structure present in histopathological images plays an important role in distinguishing malignant tissues. This paper proposes an alternate CNN network that integrates Local Binary Pattern (LBP) based texture information with CNN features. Methods: The study propagates that LBP provides the most robust rotation, and translation-invariant features in comparison with other texture feature extractors. Therefore, a formulation of LBP in context of convolution operation is presented and used in the proposed CNN network. A non-trainable fixed set binary convolutional filters representing LBP features are combined with trainable convolution filters to approximate the response of the convolution layer. A CNN architecture guided by LBP features is used to classify the histopathological images. Result: The network is trained using BreKHis datasets. The use of a fixed set of LBP filters reduces the burden of CNN by minimizing training parameters by a factor of 9. This makes it suitable for the environment with fewer resources. The proposed network obtained 96.46% of maximum accuracy with 98.51% AUC and 97% F1-score. Conclusion: LBP based texture information plays a vital role in cancer image classification. A multi-channel LBP futures fusion is used in the CNN network. The experiment results propagate that the new structure of LBP-guided CNN requires fewer training parameters preserving the capability of the CNN network’s classification accuracy.

Islamist Terrorism in the Context of Contemporary Hybrid Wars

The article deals with the present-day causes of the reproduction of Islamist terrorism. The concepts of desecularization, hybrid wars, and a system-functional approach form the methodological basis of the research. Recognizing the failure of liberal explanations of the causes of Islamist terrorism, the author criticizes the liberal methodology, which is based on an essentialist explanation of Islam and Muslim civilization and attributes a fixed set of qualities to Islam as an ontological evil, a barbarism hostile to Western civilization. The paper presents a viewpoint based on the approaches proposed by representatives of left-wing radical thought, postmodernism and neo-Marxism. It is concluded that the politicization of Islam, including its radical interpretations, is due not to the militant unchanging nature of Islam, but to the crisis of a number of Muslim societies. The Muslim worlds reaction to Western globalism is also an attempt to implement its own global political projects as a response of Islamic fundamentalism to the challenge of Western democratic fundamentalism. The author analyzes the phenomenon of hybrid wars as a form of armed violence that the Western world uses to restore order in its global empire. The connection between hybrid wars and the concept of a just war is shown, as well as the relevance of Islamist terrorism as an element of the system of hybrid wars. Islamist terrorism and counterterrorism are present in all hybrid wars waged in the Muslim world. This is manifested both in military actions on the ground, and in information warfare, as well as in virtual space. The market for terrorist and counterterrorist services inherent in hybrid wars and the place of Islamist terrorism in it are examined. Financial relations bind the participants in terrorist activities, including the customer, sponsor, mediator, organizer, informant, and performer. It is concluded that Islamist terrorism is not the activity of individual fanatics or a manifestation of the militant nature of Islam, but is produced by the conflict system of contemporary international relations.

A QR Code-Based Approach to Differentiating the Display of Augmented Reality Content

The augmented reality (AR) system requires markers to recognize and locate virtual objects on the screens of mobile devices. However, both markers and objects must be registered via the online platform in advance. In addition, an AR marker can only pair with a fixed set of virtual objects, limiting the flexibility and immediacy of changing and updating these data. This paper incorporates the quick response barcode (QR code) into the AR system to address these issues. We propose an algorithm with two vital goals, including (1) generating differentiated virtual objects for different target users by using only one QR code as the marker and (2) concealing different private authentication in QR modules by applying the error correction capability. We then demonstrate the proposed approach via a simulation of two practical scenarios, the electronic catalogs for business applications, and differentiated instructional materials for digital learning. This paper contributes to AR and QR code research and practices.

Congregations as Health Connectors – Addressing Health and Social Needs of Neighbors Through Conversations, Not Surveys

Background and Hypothesis: Social determinants of health (SDOH) directly affect health outcomes and indirectly limit access to resources needed to maintain individual health. In an effort to address the negative impacts of SDOH on the urban communities of Indianapolis, four congregations have employed Site Connectors to directly form relationships with neighbors, learn about their health journey, and connect them to resources addressing expressed needs. It is our hypothesis that through research of similar models and discussion with community partners, an optimized model for fostering relationships and assessing the health and social needs of neighbors can be developed for use by the Site Connectors. Project Methods: A search was performed for examples of health and social needs assessments utilized by established care providers, and thirteen were identified. The items within these assessments were then organized into six groups based on SDOH topic (ie. Housing, Transportation, etc.). Additionally, six interviews with community partners performing similar work were conducted utilizing a fixed set of questions. Results: It was determined that the best format for our assessment would not be a survey, as in the example assessments, but rather a visual aid resembling a concept map. This model lends itself more to the nature of relationship-building by guiding Connector-Neighbor conversations rather than dictating them, with three starter questions at the center and six offshoots covering each of the SDOH topics. A post-encounter checklist was also developed for Connectors to retroactively record priority items from their conversations. Potential Impact: It is yet to be determined whether our model will be useful in practice, as the Connectors begin their work in August. However, it is our hope that we have developed a novel format for assessing needs that more holistically addresses the impacts of SDOH through respecting the vulnerability and energy required for relationship-building.

Consolidated Advantage: New Organizational Dynamics of Wage Inequality

The two main axes of inequality in the U.S. labor market—occupation and workplace—have increasingly consolidated. In 1999, the largest share of employment at high-paying workplaces was blue-collar production workers, but by 2017 it was managers and professionals. As such, workers benefiting from a high-paying workplace are increasingly those who already benefit from membership in a high-paying occupation. Drawing on occupation-by-workplace data, we show that up to two-thirds of the rise in wage inequality since 1999 can be accounted for not by occupation or workplace inequality alone, but by this increased consolidation. Consolidation is not primarily due to outsourcing or to occupations shifting across a fixed set of workplaces. Instead, consolidation has resulted from new bases of workplace pay premiums. Workplace premiums associated with teams of professionals have increased, while premiums for previously high-paid blue-collar workers have been cut. Yet the largest source of consolidation is bifurcation in the social sector, whereby some previously low-paying but high-professional share workplaces, like hospitals and schools, have deskilled their jobs, while others have raised pay. Broadly, the results demonstrate an understudied way that organizations affect wage inequality: not by directly increasing variability in workplace or occupation premiums, but by consolidating these two sources of inequality.

Collaborative activity recognition with heterogeneous activity sets and privacy preferences

One of the major challenges in Human Activity Recognition (HAR) based on machine learning is the scarcity of labeled data. Indeed, collecting a sufficient amount of training data to build a reliable recognition problem is often prohibitive. Among the many solutions in the literature to mitigate this issue, collaborative learning is emerging as a promising direction to distribute the annotation burden over multiple users that cooperate to build a shared recognition model. One of the major issues of existing methods is that they assume a static activity model with a fixed set of target activities. In this paper, we propose a novel approach that is based on Growing When Required (GWR) neural networks. A GWR network continuously adapts itself according to the input training data, and hence it is particularly suited when the users share heterogeneous sets of activities. Like in federated learning, for the sake of privacy preservation, each user contributes to the global activity classifier by sharing personal model parameters, and not by directly sharing data. In order to further mitigate privacy threats, we implement a strategy to avoid releasing model parameters that may indirectly reveal information about activities that the user specifically marked as private. Our results on two well-known publicly available datasets show the effectiveness and the flexibility of our approach.