Task Relevance Research Articles

The Timing of Gaze Direction Perception: ERP Decoding and Task Modulation

Distinguishing the direction of another person's eye gaze is extremely important in everyday social interaction, as it provides critical information about people's attention and, therefore, intentions. The temporal dynamics of gaze processing have been investigated using event-related potentials (ERPs) recorded with electroencephalography (EEG). However, the moment at which our brain distinguishes the gaze direction (GD), irrespectively of other facial cues, remains unclear. To solve this question, the present study aimed to investigate the time course of gaze direction processing, using an ERP decoding approach, based on the combination of a support vector machine and error-correcting output codes. We recorded EEG in young healthy subjects, 32 of them performing GD detection and 34 conducting face orientation tasks. Both tasks presented 3D realistic faces with five different head and gaze orientations each: 30°, 15° to the left or right, and 0°. While the classical ERP analyses did not show clear GD effects, ERP decoding analyses revealed that discrimination of GD, irrespective of head orientation, started at 140 ms in the GD task and at 120 ms in the face orientation task. GD decoding accuracy was higher in the GD task than in the face orientation task and was the highest for the direct gaze in both tasks. These findings suggest that the decoding of brain patterns is modified by task relevance, which changes the latency and the accuracy of GD decoding.

Goal-Dependent Use of Temporal Regularities to Orient Attention under Spatial and Action Uncertainty.

The temporal regularities in our environments support the proactive dynamic anticipation of relevant events. In visual attention, one important outstanding question is whether temporal predictions must be linked to predictions about spatial locations or motor plans to facilitate behaviour. To test this, we developed a task for manipulating temporal expectations and task relevance of visual stimuli appearing within rapidly presented streams, while stimulus location and responding hand remained uncertain. Differently coloured stimuli appeared in one of two concurrent (left and right) streams with distinct temporal probability structures. Targets were defined by colour on a trial-by-trial basis and appeared equiprobably in either stream, requiring a localisation response. Across two experiments, participants were faster and more accurate at detecting temporally predictable targets compared to temporally unpredictable targets. We conclude that temporal expectations learned incidentally from temporal regularities can be called upon flexibly in a goal-driven manner to guide behaviour. Moreover, we show that visual temporal attention can facilitate performance in the absence of concomitant spatial or motor expectations in dynamically unfolding contexts.

A multi-task learning risk assessment method for the chemical process industry

Dynamic risk assessment is essential to prevent major accidents in the chemical process industry. The Prediction of accident risk categories, risk likelihood, and risk severity are three fundamental tasks for risk assessment. Single-task learning-based approaches, which merely analyze accident occurrence or quantify the likelihood and severity of an accident risk, often tend to ignore task relevance and inhibit the sharing of critical feature information. This oversight might lead to poor prediction performance. In this study, we propose a multi-task learning model, i.e., Robustly Progressive Layered Extraction (RoPLE), to systematically predict accident risk category, risk likelihood, and risk severity. The RoPLE consists of two parts: a robustly optimized pre-trained module for extracting semantic features from hazard analysis reports (unstructured data) and a progressive layered extraction module for fusing semantic features with features of leading indicators (structured data), which portray the risk state of equipment, workers, environment, and management in the chemical process industry. The progressive layered extraction module contains a multi-level extraction network and a tower network for these three tasks. The multi-level extraction network extracts higher-level shared information, while the tower network extracts task-specific feature information. Empirical experiments demonstrate the feasibility of our model on a real risk assessment dataset in Beijing’s chemical process industry. The RoPLE obtains high F1-scores of 63.1%, 79.5%, and 86.2% for the three tasks, and significantly outperforms other state-of-the-art single-task methods.

Early Electrophysiological Correlates of Perceptual Consciousness Are Affected by Both Exogenous and Endogenous Attention.

It has been proposed that visual awareness negativity (VAN), which is an early ERP component, constitutes a neural correlate of visual consciousness that is independent of perceptual and cognitive mechanisms. In the present study, we investigated whether VAN is indeed a specific marker of phenomenal awareness or rather reflects the involvement of attention. To this end, we reanalyzed data collected in a previously published EEG experiment in which awareness of visual stimuli and two aspects that define attentional involvement, namely, the inherent saliency and task relevance of a stimulus, were manipulated orthogonally. During the experimental procedure, participants (n = 41) were presented with images of faces that were backward-masked or unmasked, fearful or neutral, and defined as task-relevant targets or task-irrelevant distractors. Single-trial ERP analysis revealed that VAN was highly dependent on attentional manipulations in the early time window (140-200 msec), up to the point that the effect of awareness was not observed for attentionally irrelevant stimuli (i.e., neutral faces presented as distractors). In the late time window (200-350 msec), VAN was present in all attentional conditions, but its amplitude was significantly higher in response to fearful faces and task-relevant face images than in response to neutral ones and task-irrelevant ones, respectively. In conclusion, we demonstrate that the amplitude of VAN is highly dependent on both exogenous (stimulus saliency) and endogenous attention (task requirements). Our results challenge the view that VAN constitutes an attention-independent correlate of phenomenal awareness.

Late Positivity Correlates with Subjective Reports: Evidence from the Low-frequency and High-frequency Reporting Tasks

Identifying the neural correlates of consciousness (NCCs) is an important way to understand the fundamental nature of consciousness. By recording event-related potentials (ERPs) using EEG, researchers have found three potential electrophysiological NCCs: early positive correlate of consciousness (enhanced P1), visual awareness negativity (VAN), and late positivity (LP). However, LP may reflect post-perceptual processing associated with subjective reports rather than consciousness per se. The present experiment investigated the relationship between LP and subjective reports. We adopted two subjective reporting tasks that differed in the requirement for subjective reports. In the low-frequency reporting task, participants needed to report whether they saw the target picture in 25% of trials, whereas in the high-frequency reporting task, participants needed to report whether they saw the target picture in each trial. Behavioral results showed that the hit rates were lower and false alarm rates were higher on reporting trials in low-frequency reporting tasks than on reporting trials in high-frequency reporting tasks. Unexpectedly, VAN was larger on reporting trials in the low-frequency reporting task than on reporting trials in the high-frequency reporting task. Importantly, our ERP results showed that LP was larger on reporting trials in the high-frequency reporting task than on reporting trials in the low-frequency reporting task. Thus, our findings indicated that when the frequency of reports was increased, the task relevance of the stimuli increased, which led to larger LP amplitudes. These findings suggest that LP correlates with subjective reports.

Dissecting muscle synergies in the task space.

The muscle synergy is a guiding concept in motor control research that relies on the general notion of muscles 'working together' towards task performance. However, although the synergy concept has provided valuable insights into motor coordination, muscle interactions have not been fully characterised with respect to task performance. Here, we address this research gap by proposing a novel perspective to the muscle synergy that assigns specific functional roles to muscle couplings by characterising their task-relevance. Our novel perspective provides nuance to the muscle synergy concept, demonstrating how muscular interactions can 'work together' in different ways: (1) irrespective of the task at hand but also (2) redundantly or (3) complementarily towards common task-goals. To establish this perspective, we leverage information- and network-theory and dimensionality reduction methods to include discrete and continuous task parameters directly during muscle synergy extraction. Specifically, we introduce co-information as a measure of the task-relevance of muscle interactions and use it to categorise such interactions as task-irrelevant (present across tasks), redundant (shared task information), or synergistic (different task information). To demonstrate these types of interactions in real data, we firstly apply the framework in a simple way, revealing its added functional and physiological relevance with respect to current approaches. We then apply the framework to large-scale datasets and extract generalizable and scale-invariant representations consisting of subnetworks of synchronised muscle couplings and distinct temporal patterns. The representations effectively capture the functional interplay between task end-goals and biomechanical affordances and the concurrent processing of functionally similar and complementary task information. The proposed framework unifies the capabilities of current approaches in capturing distinct motor features while providing novel insights and research opportunities through a nuanced perspective to the muscle synergy.

Isolating Neural Signatures of Conscious Speech Perception with a No-Report Sine-Wave Speech Paradigm.

Identifying neural correlates of conscious perception is a fundamental endeavor of cognitive neuroscience. Most studies so far have focused on visual awareness along with trial-by-trial reports of task-relevant stimuli, which can confound neural measures of perceptual awareness with postperceptual processing. Here, we used a three-phase sine-wave speech paradigm that dissociated between conscious speech perception and task relevance while recording EEG in humans of both sexes. Compared with tokens perceived as noise, physically identical sine-wave speech tokens that were perceived as speech elicited a left-lateralized, near-vertex negativity, which we interpret as a phonological version of a perceptual awareness negativity. This response appeared between 200 and 300 ms after token onset and was not present for frequency-flipped control tokens that were never perceived as speech. In contrast, the P3b elicited by task-irrelevant tokens did not significantly differ when the tokens were perceived as speech versus noise and was only enhanced for tokens that were both perceived as speech and relevant to the task. Our results extend the findings from previous studies on visual awareness and speech perception and suggest that correlates of conscious perception, across types of conscious content, are most likely to be found in midlatency negative-going brain responses in content-specific sensory areas.

Event-induced modulation of aperiodic background EEG: Attention-dependent and age-related shifts in E:I balance, and their consequences for behavior

Abstract The broadband shape of the EEG spectrum, summarized using the slope of a 1/fx function, is thought to reflect the balance between excitation and inhibition in cortical regions (E:I balance). This balance is an important characteristic of neural circuits and could inform studies of aging, as older adults show a relative deficit in inhibitory activity. Thus far, no studies have leveraged the event-related temporal dynamics of 1/fx activity to better understand the phases of information processing, especially in the context of aging. Here, for the first time, we examined variations of this activity during the foreperiod of a cued flanker task in younger (YA) and older adults (OA), with picture cues varying in task relevance, relative novelty, and valence. We report a biphasic change in the spectral slope after cue presentation, independent of cue-elicited event-related potentials (ERPs), with an initial period of steeper slope (indicating cortical inhibition, similar in YA and OA) followed by a flattening (indicating cortical excitation, especially in OA). The reduction in slope steepness was associated with lower performance and greater congruency costs in the flanker task. Finally, more novel cues reduced the shift towards excitation in OA, partly restoring their E:I balance, and diminishing congruency costs. These findings demonstrate that the broadband shape of the EEG spectrum varies dynamically in a manner that is predictive of subsequent behavior. They also expand our understanding of how neural communication shapes cognition in YA and OA and has implications for neuroscientific models of cognitive processing and age-related cognitive decline.

Bidirectional Gated Recurrent Unit with Glove Embedding and Attention Mechanism for Movie Review Classification

During the current Internet era, enormous volume of structured and unstructured textual data is generated and exchanged online that made the text classification more crucial. With the massive amount of user-generated content on social media platforms, businesses and organizations can leverage sentiment analysis to understand customer opinions and attitudes towards the products, services, or brand. Text classification is a natural language processing task in which a machine learning model is trained to categorize the text into predefined classes or categories. Due to the advent of promising Deep Learning techniques, intelligent and accurate text classification system can be developed. In movie review classification, algorithms are used to automatically categorize movie reviews into positive, negative, or sometimes neutral sentiments based on the opinions expressed in the text. The goal is to automate the process of determining whether a review reflects a positive or negative sentiment, helping users quickly understand the overall reception of a movie. This work proposes an automated movie review classification system based on Bidirectional Gated Recurrent Unit encoder with attention module. This method encodes the words using GloVe word embedding technique. Context information is captured in an efficient manner and the encoding is more reliable. Bidirectional GRU layer captures long-term dependencies in sentences, followed by a custom Attention layer that assigns higher weights to key components. The attention mechanism allows the model to selectively focus on important parts of the input sequence, by dynamically determining attention weights based on task relevance. The model was individually trained and evaluated using popular benchmark IMDb dataset. Experimental studies have proved the better performance of proposed movie review classification system that obtained accuracy of 98%.

Subliminal Cueing in Visual Attention: Top-down and Bottomup Processing

Exogenous subliminal cues have been shown to have both top-down and bottom-up effects. There are many research studies have confirmed that the effect of top-down variables dependent on cue characteristics. In response, other studies have attempted to exhibit ‘purely’ stimulus-driven attention capture. The superiority of attention capture effects and the level of superiority in top-down cues have not been evidenced previously. The present study attempted to observe differential effects of cues with different levels of task relevancy, in both valid and invalid cues. As an addendum, conditions with valid cues and incorrect feature match allowed for the exploration of same-location costs in subliminal cues. The results indicated attentional capture effects of all valid cues irrespective of the level of task relevance in subliminal conditions. The cues with the highest task relevancy led to the most attention capture. There were no inhibitory effects of invalid cues on attention capture in subliminal cues. The study reiterates the possibility of attentional mechanisms in same-location costs implicating processing outside of memory. Further, facilitation and inhibition of attention capture in subliminal cues may have two independent mechanisms. The inhibitory mechanism may be contingent to conscious awareness of cues. The result of the study shows how attention can be exogenously oriented without conscious awareness and what kinds of cues might be most effective. These findings can be useful in educational and training settings and other settings where sustained attention is crucial.

The effect of task relevance on serial dependence in numerosity

The effect of task relevance on serial dependence in numerosity

The effect of multisensory distraction on working memory: A role for task relevance?

There is growing recognition that working memory and selective attention are highly related. However, a key function of selective attention-ignoring distractors-is much less understood in the domain of working memory. In the attention domain, it is now clear that distractors' task relevance and stimulation of multiple senses at a time (i.e., being multisensory), affect how much such information can distract from the main task, and that load modulates these effects. Here, we examined the effects of the task relevance and multisensory nature of distractors on working memory performance under high and low memory load, aiming to clarify whether distracting information similarly affects selective attention performance and working memory performance. We proposed a multiexperiment research plan involving up to three consecutive experiments, based on an initial online study (Experiment 0) with fully task-irrelevant distractors. There, we found conclusive evidence against a difference in how unisensory and multisensory distractors affected working memory performance. The next study, Experiment 1, replicated these results. However, when distractors were made partly task relevant in the subsequent Experiment 2d, multisensory distractors disrupted working memory performance more than unisensory distractors on average. However, closer nonpreregistered inspection revealed that multisensory distractors were actually only more disruptive than auditory distractors, and similarly as disruptive as visual distractors. Thus, overall, there was no strong evidence for multisensory distractors being more disruptive to working memory performance than unisensory distractors. Taken together, these experiments constitute a novel and detailed investigation of the impact of distracting information on working memory performance. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Supplemental Material for The Effect of Multisensory Distraction on Working Memory: A Role for Task Relevance?

Supplemental Material for The Effect of Multisensory Distraction on Working Memory: A Role for Task Relevance?

Stimulus awareness is associated with secondary somatosensory cortex activation in an inattentional numbness paradigm

While inattentional blindness and deafness studies have revealed neural correlates of consciousness (NCC) without the confound of task relevance in the visual and auditory modality, comparable studies for the somatosensory modality are lacking. Here, we investigated NCC using functional magnetic resonance imaging (fMRI) in an inattentional numbness paradigm. Participants (N = 44) received weak electrical stimulation on the left hand while solving a demanding visual task. Half of the participants were informed that task-irrelevant weak tactile stimuli above the detection threshold would be applied during the experiment, while the other half expected stimuli below the detection threshold. Unexpected awareness assessments after the experiment revealed that altogether 10 participants did not consciously perceive the somatosensory stimuli during the visual task. Awareness was not significantly modulated by prior information. The fMRI data show that awareness of stimuli led to increased activation in the contralateral secondary somatosensory cortex. We found no significant effects of stimulus awareness in the primary somatosensory cortex or frontoparietal areas. Thus, our results support the hypothesis that somatosensory stimulus awareness is mainly based on activation in higher areas of the somatosensory cortex and does not require strong activation in extended anterior or posterior networks, which is usually seen when perceived stimuli are task-relevant.

Relevance and the Role of Labels in Categorization.

Language has been shown to influence the ability to form categories. Nevertheless, in most prior work, the effects of language could have been bolstered by the fact that linguistic labels were introduced by the experimenter prior to the categorization task in ways that could have highlighted their relevance for the task. Here, we compared the potency of labels to that of other non-linguistic cues on how people categorized novel, perceptually ambiguous natural kinds (e.g., flowers or birds). Importantly, we varied whether these cues were explicitly presented as relevant to the categorization task. In Experiment 1, we compared labels, numbers, and symbols: One group of participants was told to pay attention to these cues because they would be helpful (Relevant condition), a second group was told that the cues were irrelevant and should be ignored (Irrelevant condition), and a third group was told nothing about the cues (Neutral condition). Even though task relevance affected overall reliance on cues during categorization, participants were more likely to use labels to determine category boundaries, compared to numbers or symbols. In Experiments 2 and 3, we replicated and fine-tuned the advantage of labels in more stringent categorization tasks. These results offer novel evidence for the position that labels offer unique indications of category membership, compared to non-linguistic cues.

APGVAE: Adaptive disentangled representation learning with the graph-based structure information

Neural networks are used to learn task-oriented high-level representations in an end-to-end manner by building a multi-layer neural network. Generation models have developed rapidly with the emergence of deep neural networks. But it still has problems with the insufficient authenticity of generated images, the deficiency of diversity, consistency, and unexplainability in the generation process. Disentangled representation is an effective method to learn a high-level feature representation and realize the interpretability of deep neural networks. We propose a general disentangled representation learning network with variational autoencoder network as the basic framework for the image generation process. The graph-based structure of the priors is embedded in the last module of the deep encoder network to build the feature spaces by the class, task-oriented, and task-unrelated information respectively. Meanwhile the priors should be adaptively modified with the task relevance of a generated image. And the semi-supervised learning is further involved in the disentangled representation network framework to reduce the requirements of label and extend the majority of feature space under the task-unrelated feature assumption. Experimental results show that the proposed method is efficient for various types of images and has a good potential for further research and development.

Automatizing Sight Reading: Contingency Proportion and Task Relevance in the Music Contingency Learning Procedure

Learning to read and play music written in standard notation, termed sight reading, is an important yet difficult aspect of early music education. However, the music contingency learning procedure produces rapid and robust early learning of the motor execution associated with note positions. In this task, nonmusicians identify a note name (e.g., “do”) written inside a note in one of the vertical positions of the musical staff with a keyboard response. Each note position is presented frequently with the matching (congruent) note name and rarely with the incongruent note names. The present work further explores this novel learning paradigm. In Experiment 1, we manipulated the proportion of congruent trials from 50 to 100%. The contingency effect, along with contingency awareness (i.e., verbalizable knowledge of note meanings), increased with a stronger contingency manipulation. In Experiment 2, half of the participants responded to the note positions (instead of the note names) with a keyboard response. A learning effect was also observed for this task, though contingency awareness was reduced in this group. These results shed more light on the properties of incidental music learning and further suggest more ideal parameters for future practical applications to supplement traditional instruction in real-world music education.

Effects of sensory modality and task relevance on omitted stimulus potentials.

Omitted stimulus potentials (OSPs) occur when a sensory stimulus is unexpectedly omitted. They are thought to reflect predictions about upcoming sensory events. The present study examined how OSPs differ across the sensory modalities of predicted stimuli. Twenty-nine university students were asked to press a mouse button at a regular interval of 1-2s, which was immediately followed by either a visual or auditory stimulus in different blocks. The stimuli were sometimes omitted (p = 0.2), to which event-related potentials (ERPs) were recorded. The results showed that stimulus omissions in both modalities elicited ERP waveforms consisting of three components, oN1, oN2, and oP3. The peak latencies of these components were shorter in the auditory modality than in the visual modality. The amplitudes of OSPs were larger when participants were told that the omission indicated their poor performance (i.e., they pressed a button at an irregular interval) than when it was irrelevant to their performance. These findings suggest that OSPs occur from around 100ms in a modality-specific manner and increase in amplitude depending on the task relevance of stimulus omissions.

ADDNS: An asymmetric dual deep network with sharing mechanism for medical image fusion of CT and MR-T2

Medical images with different modalities have different semantic characteristics. Medical image fusion aiming to promotion of the visual quality and practical value has become important in medical diagnostics. However, the previous methods do not fully represent semantic and visual features, and the model generalization ability needs to be improved. Furthermore, the brightness-stacking phenomenon is easy to occur during the fusion process. In this paper, we propose an asymmetric dual deep network with sharing mechanism (ADDNS) for medical image fusion. In our asymmetric model-level dual framework, primal Unet part learns to fuse medical images of different modality into a fusion image, while dual Unet part learns to invert the fusion task for multi-modal image reconstruction. This asymmetry of network settings not only enables the ADDNS to fully extract semantic and visual features, but also reduces the model complexity and accelerates the convergence. Furthermore, the sharing mechanism designed according to task relevance also reduces the model complexity and improves the generalization ability of our model. In the end, we use the intermediate supervision method to minimize the difference between fusion image and source images so as to prevent the brightness-stacking problem. Experimental results show that our algorithm achieves better results on both quantitative and qualitative experiments than several state-of-the-art methods.

Advancing hearing research with the NAL ecologically momentary assessment platform for real-world insights

Ecological Momentary Assessment (EMA) is an invaluable tool for assessing people's behaviours and experiences in their natural surroundings. We have developed a smartphone-based EMA tool, called NEMA, that has contributed to over 10 research studies by providing meaningful real-world insights into how individuals with hearing loss interact with various hearing device technologies. By providing real-world perspectives that complement traditional lab-based tests, NEMA offers a more comprehensive understanding of everyday hearing and communication challenges. Its advanced features include cloud-connectivity, customisable surveys, cross-device acoustic feature measurements from the environment, and a back-end platform enabling real-time data monitoring and visualization. To derive vital insights from this rich data stream, advanced analytical methodologies are applied. However, there exist limitations with smartphone-based EMAs that can influence their overall effectiveness. We identify these constraints and discuss strategies to alleviate participant burden and increase task relevance. These initiatives aim to enhance compliance and diversify data collected. Current work involves integrating smartwatches and additional sensor technology into our next generation NEMA platform. This progression aims to facilitate a more comprehensive capture of diverse data sets, ultimately leading to a deeper understanding of individual auditory experiences.