Dimensional Labeling Research Articles

Examining Neural Dynamics During Dimensional Label Comprehension and Production as a Function of Dimensional Attention

ABSTRACT Previous research suggests that children’s ability to label visual features (e.g. “red”) and dimensions (e.g. “color”) impacts attention to visual dimensions. The goal of this study is to investigate variations in the quality of the neural system supporting dimensional label comprehension and production in relation to children’s dimensional attention skills. The current study recruited 3- to 4-year-old from the East Tennessee area to complete dimensional label tasks for color labels while recording functional near-infrared spectroscopy (fNIRS) in order to measure hemodynamic changes in left frontal, left parietal, and left temporal cortices previously implicated in dimensional attention. Different aspects of dimensional attention were evaluated using the dimensional priming task (measuring attentional stability), the dimensional change card sort (DCCS) task (measuring flexible dimensional attention), the triad classification task (measuring selective attention), and the matching task (measuring dimensional understanding). Dimensional label learning was measured using a color Production and color Comprehension task. Results indicated that the temporal cortex was activated during the dimensional label tasks. Further, we found that activation in parietal cortex during the dimensional label tasks increased with better performance on the dimensional attention tasks. On the other hand, activation in the temporal cortex and ventrolateral prefrontal cortex during the dimensional label tasks decreased with better performance on the dimensional attention tasks. These results suggest that dimensional attention skills can reveal variations in the quality of neural representations supporting dimensional label learning.

Dimensional label learning contributes to the development of executive functions

A key to understanding how the brain develops is to understand how learning can change brain function. One index of learning that takes place in early childhood involves the comprehension and production of labels describing the shape and color features of objects, a process known as dimensional label learning (DLL). DLL requires integrating auditory and visual stimuli to form a system of mappings that link label representations (e.g. “red” and “color”) and visual feature representations (e.g. “red” and the hue red). Children gain expertise with these labels between the ages of 2 and 5 years, and at the same time they begin to demonstrate skills in using labels to guide cognitive function in other domains. For example, one of the hallmark measures of executive function development requires children to use verbally instructed rules to guide attention to visual dimensions. The broader impact of DLL, however, has not yet been explored. Here, we examine how the neural processes associated with the comprehension and production of labels for visual features predicts later performance on executive function tasks. Specifically, we show that left frontal cortex is activated during comprehension and production tasks at 33 months of age. Moreover, we find that neural activation in this region during label production at 33 months is associated with dimensional attention, but not spatial selective attention, at 45 months. These results shed new light on the role of label learning in developmental changes in brain and behavior. Moreover, these data suggest that dimensional label learning generalizes beyond the learned information to influence other aspects of cognition. We anticipate that these results may serve as a starting point for future work to implement label training as an intervention to influence later cognition.

Subjective Evaluation of Basic Emotions from Audio–Visual Data

Understanding of the perception of emotions or affective states in humans is important to develop emotion-aware systems that work in realistic scenarios. In this paper, the perception of emotions in naturalistic human interaction (audio–visual data) is studied using perceptual evaluation. For this purpose, a naturalistic audio–visual emotion database collected from TV broadcasts such as soap-operas and movies, called the IIIT-H Audio–Visual Emotion (IIIT-H AVE) database, is used. The database consists of audio-alone, video-alone, and audio–visual data in English. Using data of all three modes, perceptual tests are conducted for four basic emotions (angry, happy, neutral, and sad) based on category labeling and for two dimensions, namely arousal (active or passive) and valence (positive or negative), based on dimensional labeling. The results indicated that the participants’ perception of emotions was remarkably different between the audio-alone, video-alone, and audio–video data. This finding emphasizes the importance of emotion-specific features compared to commonly used features in the development of emotion-aware systems.

Abstract 13912: ECG-DualNet: A MultiLabel Model for Automated Diagnosis on Both Electrocardiographic Signals and Printed 12-lead Electrocardiograms

Background: Electrocardiograms (ECGs) are frequently stored as printed images, but tools developed for artificial intelligence (AI)-based automated diagnosis and patient phenotyping exclusively rely on signal data, limiting their application. Methods: We developed a model that simultaneously incorporates both signal and image-based ECG data in clinical diagnosis and patient phenotyping. We used all 12-lead ECGs from a large deidentified database from Brazil, with signals transformed to a standard 12-lead ECG image. Both ECG signals and images were randomly subset into training, validation and test sets (90%-5%-5%). We used 6 physician-defined clinical labels spanning rhythm and conduction disorders as labels (atrial fibrillation [AF], sinus tachy/bradycardia [ST/SB], 1 st degree AV block [1dAVB], RBBB, LBBB). In addition, ECG-based identification of patient sex was assessed as a higher dimensional label. Convolutional Neural Networks were constructed with custom Inception net architecture for signals and EfficientNet-B3 for images. Model performance was evaluated in the holdout test set and US-based external validation dataset, PTB-XL. Results: The derivation dataset included 2,228,236 ECGs in 1,506,112 individuals (mean age 54y, 60% female). The ECG-DualNet model successfully labeled all 6 clinical diagnoses and patient gender with a high model performance (AUC of .81-.99 for signal and .92-.99 for images, F1 scores, .31-.78, and .49-82) in the 111,412 held-out test ECGs (Fig A-B). In external validation, with 21,785 ECGs, the model had excellent performance on both ECG signals and images (AUCs .75-.99 and .89-.99, F1 scores, .25-82, and .44-.86) (Fig C-D). Conclusion: A novel multilabel dual signal and image-based model achieved high performance in both clinical diagnosis and higher dimensional phenotypic labels across distinct ECG sources, with the ability to standardize automated diagnosis and patient phenotyping across ECG data formats.

Not all labels develop equally: The role of labels in guiding attention to dimensions

The emergence of cognitive flexibility is a central aspect of cognitive development during early childhood. Cognitive flexibility is often probed using verbal rules to instruct behavior. In this study, the types of labels that were provided during instruction were manipulated. In one condition, children were instructed in the standard manner with dimensional labels (e.g., “shape”) and featural labels (e.g., “star”). In a second condition, children were provided only with dimensional labels. When switching to color, 4-year-olds performed equally well regardless of the type of instruction. However, when switching to shape, children perseverated at a significantly higher rate when only dimensional labels were provided. These results suggest that children’s understanding of labels is a critical aspect of developing cognitive flexibility and that their understanding of the labels “shape” and “color” are different.

Joint Ranking SVM and Binary Relevance with robust Low-rank learning for multi-label classification

Multi-label classification studies the task where each example belongs to multiple labels simultaneously. As a representative method, Ranking Support Vector Machine (Rank-SVM) aims to minimize the Ranking Loss and can also mitigate the negative influence of the class-imbalance issue. However, due to its stacking-style way for thresholding, it may suffer error accumulation and thus reduces the final classification performance. Binary Relevance (BR) is another typical method, which aims to minimize the Hamming Loss and only needs one-step learning. Nevertheless, it might have the class-imbalance issue and does not take into account label correlations. To address the above issues, we propose a novel multi-label classification model, which joints Ranking support vector machine and Binary Relevance with robust Low-rank learning (RBRL). RBRL inherits the ranking loss minimization advantages of Rank-SVM, and thus overcomes the disadvantages of BR suffering the class-imbalance issue and ignoring the label correlations. Meanwhile, it utilizes the hamming loss minimization and one-step learning advantages of BR, and thus tackles the disadvantages of Rank-SVM including another thresholding learning step. Besides, a low-rank constraint is utilized to further exploit high-order label correlations under the assumption of low dimensional label space. Furthermore, to achieve nonlinear multi-label classifiers, we derive the kernelization RBRL. Two accelerated proximal gradient methods (APG) are used to solve the optimization problems efficiently. Extensive comparative experiments with several state-of-the-art methods illustrate a highly competitive or superior performance of our method RBRL.

Dimensional attention as a mechanism of executive function: Integrating flexibility, selectivity, and stability

In this report, we present a neural process model that explains visual dimensional attention and changes in visual dimensional attention over development. The model is composed of an object representation system that binds visual features such as shape and color to spatial locations and a label learning system that associates labels such as “color” or “shape” with visual features. We have previously demonstrated that this model explains the development of flexible dimensional attention in a task that requires children to switch between shape and color rules for sorting cards. In the model, the development of flexible dimensional attention is a product of strengthening associations between labels and features. In this report, we generalize this model to also explain development of stable and selective dimensional attention. Specifically, we use the model to explain a previously reported developmental association between flexible dimensional attention and stable dimensional attention. Moreover, we generate predictions regarding developmental associations between flexible and selective dimensional attention. Results from an experiment with 3- and 4-year-olds supported model predictions: children who demonstrated flexibility also demonstrated higher levels of selectivity. Thus, the model provides a framework that integrates various functions of dimensional attention, including implicit and explicit functions, over development. This model also provides new avenues of research aimed at uncovering how cognitive functions such as dimensional attention emerge from the interaction between neural dynamics and task structure, as well as understanding how learning dimensional labels creates changes in dimensional attention, brain activation, and neural connectivity.

ACCELERATED FITTING OF STELLAR SPECTRA

ABSTRACT Stellar spectra are often modeled and fitted by interpolating within a rectilinear grid of synthetic spectra to derive the stars’ labels: stellar parameters and elemental abundances. However, the number of synthetic spectra needed for a rectilinear grid grows exponentially with the label space dimensions, precluding the simultaneous and self-consistent fitting of more than a few elemental abundances. Shortcuts such as fitting subsets of labels separately can introduce unknown systematics and do not produce correct error covariances in the derived labels. In this paper we present a new approach—Convex Hull Adaptive Tessellation (chat)—which includes several new ideas for inexpensively generating a sufficient stellar synthetic library, using linear algebra and the concept of an adaptive, data-driven grid. A convex hull approximates the region where the data lie in the label space. A variety of tests with mock data sets demonstrate that chat can reduce the number of required synthetic model calculations by three orders of magnitude in an eight-dimensional label space. The reduction will be even larger for higher dimensional label spaces. In chat the computational effort increases only linearly with the number of labels that are fit simultaneously. Around each of these grid points in the label space an approximate synthetic spectrum can be generated through linear expansion using a set of “gradient spectra” that represent flux derivatives at every wavelength point with respect to all labels. These techniques provide new opportunities to fit the full stellar spectra from large surveys with 15–30 labels simultaneously.

Tangential Approximation of Surfaces

In the Computer Vision community it is a common belief that higher order smoothness, such as curvature, should be modeled using higher order interactions. For example, 2nd order derivatives for deformable (active) contours are represented by triple cliques. Similarly, the 2nd order regularization methods in stereo predominantly use MRF models with scalar (1D) disparity labels and triple clique interactions. In this paper we give an overview of an energy minimization framework for tangential approximation of surfaces developed in [21, 22]. The framework uses higher dimensional labels to encode second order smoothness with pairwise interactions. Hence, many generic optimization algorithms (e.g. message passing, graph cut, etc.) can be used to optimize the proposed regularization functionals. The accuracy of our approach for representing curvature is demonstrated by theoretical and empirical results on real data sets from multi-view reconstruction and stereo.

Tight Convex Relaxations for Vector-Valued Labeling

Multilabel problems are of fundamental importance in computer vision and image analysis. Yet, finding global minima of the associated energies is typically a hard computational challenge. Recently, progress has been made by reverting to spatially continuous formulations of respective problems and solving the arising convex relaxation globally. In practice this leads to solutions which are either optimal or within an a posteriori bound of the optimum. Unfortunately, in previous methods, both run time and memory requirements scale linearly in the total number of labels, making these methods very inefficient and often not applicable to problems with higher dimensional label spaces. In this paper, we propose a reduction technique for the case that the label space is a continuous product space and the regularizer is separable, i.e., a sum of regularizers for each dimension of the label space. In typical real-world labeling problems, the resulting convex relaxation requires orders of magnitude less memory and c...

Discriminant Analysis for Fast Multiclass Data Classification Through Regularized Kernel Function Approximation

In this brief we have proposed the multiclass data classification by computationally inexpensive discriminant analysis through vector-valued regularized kernel function approximation (VVRKFA). VVRKFA being an extension of fast regularized kernel function approximation (FRKFA), provides the vector-valued response at single step. The VVRKFA finds a linear operator and a bias vector by using a reduced kernel that maps a pattern from feature space into the low dimensional label space. The classification of patterns is carried out in this low dimensional label subspace. A test pattern is classified depending on its proximity to class centroids. The effectiveness of the proposed method is experimentally verified and compared with multiclass support vector machine (SVM) on several benchmark data sets as well as on gene microarray data for multi-category cancer classification. The results indicate the significant improvement in both training and testing time compared to that of multiclass SVM with comparable testing accuracy principally in large data sets. Experiments in this brief also serve as comparison of performance of VVRKFA with stratified random sampling and sub-sampling.

Optimal Elections in Labeled Hypercubes

We study the message complexity of theElectionProblem in hypercube networks, when the processors have a “Sense of Direction,” i.e., the capability to distinguish between adjacent communication links according to some globally consistent scheme. We present two models of Sense of Direction, which differ regarding the way the labeling of the links in the network is done: either by matching based on dimensions or by distance along a Hamiltonian cycle. In the dimension model, we give an optimal linear algorithm which uses the natural dimensional labeling of the communication links. We prove that, in the distance-based case, the graph symmetry of the hypercube is broken and, thus, the leader Election does not require a global maximum-finding algorithm:O(1) messages suffice to select the leader, whereas the Θ(N) messages are required only for the final broadcasting. Finally, we study the communication cost of changing one orientation labeling to the other and prove thatO(N) messages suffice.

Analysis of proteins synthesized by fibroblasts from patients with cystic fibrosis by two-dimensional gel electrophoresis and double label autoradiography.

Mucoviscidosis, the most frequently lethal genetic syndrome of Caucasian population, is a recessive disease with multiple tissue involvement. Although the major pathological changes are observed in lungs and pancreas, abnormalities have also been detected in several other exocrine glands. For many reasons, such as the ready availability of tissue material, the absence of secondary changes and the potential for prenatal diagnosis, cultured skin fibroblasts could be the tissue of choice to search for the primary defect. Several abnormalities have been reported in CF fibroblasts, suggesting that the genetic abnormality is expressed in these cells. To search for potentially mutant protein(s) we have compared the protein composition of normal and CF fibroblasts by two dimensional gel electrophoresis and double-labeling autoradiography using 35S and 75Se methionine as tracer. The results demonstrate the power of the method; however, we have not found one protein spot consistently missing in CF cells. Possible reasons for the absence of a single common identifiable defect are discussed.

The effects of labeling dimensional values on setting differences in shift performance of kindergarten children

A reversal-shift paradigm was used to study the effects of dimensional labeling on the performance of 270 kindergarten children. Task 1 was a simultaneous discrimination problem in which form, size, and color were redundant relevant dimensions, In the shift task, one of the dimensions remained relevant with reward contingencies reversed, and the other two dimensions were made irrelevant. In both tasks, the children named the stimulus object to be chosen, prior to choosing, in terms of one of the three sets of dimensionnal values. Shift performance was predicted to be superior in groups required to name relevant dimensional values in comparison with groups required to name irrelevant dimensional values. Moreover shift performance was predicted to differ, within groups, for four different types of stimulus settings. Fmally, differences in performance on the settings were predicted to be larger for the irrelevant-labeling groups than for the relevant-labelmg groups. The results were found to be in good agreement with these predictions. The findings were discussed in relation to the subproblem analysis proposed by Tighe and Tighe (1972).