Low-level Tasks Research Articles

Exploring the categorical nature of colour perception: Insights from artificial networks

The electromagnetic spectrum of light from a rainbow is a continuous signal, yet we perceive it vividly in several distinct colour categories. The origins and underlying mechanisms of this phenomenon remain partly unexplained. We investigate categorical colour perception in artificial neural networks (ANNs) using the odd-one-out paradigm. In the first experiment, we compared unimodal vision networks (e.g., ImageNet object recognition) to multimodal vision-language models (e.g., CLIP text-image matching). Our results show that vision networks predict a significant portion of human data (approximately 80%), while vision-language models account for the remaining unexplained data, even in non-linguistic experiments. These findings suggest that categorical colour perception is a language-independent representation, though it is partly shaped by linguistic colour terms during its development. In the second experiment, we explored how the visual task influences the colour categories of an ANN by examining twenty-four Taskonomy networks. Our results indicate that human-like colour categories are task-dependent, predominantly emerging in semantic and 3D tasks, with a notable absence in low-level tasks. To explain this difference, we analysed kernel responses before the winner-takes-all stage, observing that networks with mismatching colour categories may still align in underlying continuous representations. Our findings quantify the dual influence of visual signals and linguistic factors in categorical colour perception and demonstrate the task-dependent nature of this phenomenon, suggesting that categorical colour perception emerges to facilitate certain visual tasks.

A survey on pragmatic processing techniques

Pragmatics, situated in the domains of linguistics and computational linguistics, explores the influence of context on language interpretation, extending beyond the literal meaning of expressions. It constitutes a fundamental element for natural language understanding in machine intelligence. With the advancement of large language models, the research focus in natural language processing has predominantly shifted toward high-level task processing, inadvertently downplaying the importance of foundational pragmatic processing tasks. Nevertheless, pragmatics serves as a crucial medium for unraveling human language cognition. The exploration of pragmatic processing stands as a pivotal facet in realizing linguistic intelligence. This survey encompasses important pragmatic processing techniques for subjective and emotive tasks, such as personality recognition, sarcasm detection, metaphor understanding, aspect extraction, and sentiment polarity detection. It spans theoretical research, the forefront of pragmatic processing techniques, and downstream applications, aiming to highlight the significance of these low-level tasks in advancing natural language understanding and linguistic intelligence.

A novel multiscale cGAN approach for enhanced salient object detection in single haze images

In computer vision, image dehazing is a low-level task that employs algorithms to analyze and remove haze from images, resulting in haze-free visuals. The aim of Salient Object Detection (SOD) is to locate the most visually prominent areas in images. However, most SOD techniques applied to visible images struggle in complex scenarios characterized by similarities between the foreground and background, cluttered backgrounds, adverse weather conditions, and low lighting. Identifying objects in hazy images is challenging due to the degradation of visibility caused by atmospheric conditions, leading to diminished visibility and reduced contrast. This paper introduces an innovative approach called Dehaze-SOD, a unique integrated model that addresses two vital tasks: dehazing and salient object detection. The key novelty of Dehaze-SOD lies in its dual functionality, seamlessly integrating dehazing and salient object identification into a unified framework. This is achieved using a conditional Generative Adversarial Network (cGAN) comprising two distinct subnetworks: one for image dehazing and another for salient object detection. The first module, designed with residual blocks, Dark Channel Prior (DCP), total variation, and the multiscale Retinex algorithm, processes the input hazy images. The second module employs an enhanced EfficientNet architecture with added attention mechanisms and pixel-wise refinement to further improve the dehazing process. The outputs from these subnetworks are combined to produce dehazed images, which are then fed into our proposed encoder–decoder framework for salient object detection. The cGAN is trained with two modules working together: the generator aims to produce haze-free images, whereas the discriminator distinguishes between the generated haze-free images and real haze-free images. Dehaze-SOD demonstrates superior performance compared to state-of-the-art dehazing methods in terms of color fidelity, visibility enhancement, and haze removal. The proposed method effectively produces high-quality, haze-free images from various hazy inputs and accurately detects salient objects within them. This makes Dehaze-SOD a promising tool for improving salient object detection in challenging hazy conditions. The effectiveness of our approach has been validated using benchmark evaluation metrics such as mean absolute error (MAE), peak signal-to-noise ratio (PSNR), and structural similarity index measure (SSIM).

Emotion Processing and Its Relationship to Social Functioning and Symptoms in Psychotic Disorder: A Systematic Review and Meta-analysis.

Emotion processing (EP) is impaired in individuals with psychosis and associated with social functioning; however, it is unclear how symptoms fit into this relationship. The aim of this systematic review and meta-analysis was to examine interrelationships between EP, symptoms, and social functioning, test whether different symptom domains mediate the relationship between EP and social functioning, and examine the moderating effects of illness stage and EP task type. MEDLINE, Embase, and PsycINFO databases were searched for studies that included individuals with psychosis and reported correlations between EP, symptom domains (positive, negative, depressive, and disorganization), and social functioning. Random effects meta-analyses determined the strength of correlations, and subgroup analyses included illness stage and EP task type (lower- vs higher-level processing). Meta-analytic structural equation models tested whether symptom domains mediated the relationship between EP and social functioning. There was a small relationship (r = .18) between EP and social functioning. Positive, negative, and disorganization symptoms mediated this relationship, although indirect effects were small. Higher-level EP tasks were more strongly associated with negative symptoms than lower-level tasks. Relationships between EP and both social functioning and positive symptoms were smaller in the first episode of psychosis than in established illness. The mediating relationship suggests that EP not only influences social dysfunction directly but contributes to negative and disorganization symptoms, which in turn impair social functioning. This pathway suggests that targeting negative and disorganization symptoms may ultimately improve social outcomes for individuals with psychosis. Future research, particularly in early psychosis, is needed to determine other factors impacting these interrelationships.

Enhancing global sensitivity and uncertainty quantification in medical image reconstruction with Monte Carlo arbitrary-masked mamba

Deep learning has been extensively applied in medical image reconstruction, where Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs) represent the predominant paradigms, each possessing distinct advantages and inherent limitations: CNNs exhibit linear complexity with local sensitivity, whereas ViTs demonstrate quadratic complexity with global sensitivity. The emerging Mamba has shown superiority in learning visual representation, which combines the advantages of linear scalability and global sensitivity. In this study, we introduce MambaMIR, an Arbitrary-Masked Mamba-based model with wavelet decomposition for joint medical image reconstruction and uncertainty estimation. A novel Arbitrary Scan Masking (ASM) mechanism “masks out” redundant information to introduce randomness for further uncertainty estimation. Compared to the commonly used Monte Carlo (MC) dropout, our proposed MC-ASM provides an uncertainty map without the need for hyperparameter tuning and mitigates the performance drop typically observed when applying dropout to low-level tasks. For further texture preservation and better perceptual quality, we employ the wavelet transformation into MambaMIR and explore its variant based on the Generative Adversarial Network, namely MambaMIR-GAN. Comprehensive experiments have been conducted for multiple representative medical image reconstruction tasks, demonstrating that the proposed MambaMIR and MambaMIR-GAN outperform other baseline and state-of-the-art methods in different reconstruction tasks, where MambaMIR achieves the best reconstruction fidelity and MambaMIR-GAN has the best perceptual quality. In addition, our MC-ASM provides uncertainty maps as an additional tool for clinicians, while mitigating the typical performance drop caused by the commonly used dropout.

Toward Constructing Frameworks for Task- and Design-Optimized Visualizations.

Visualization is crucial to augment and enhance human understanding and decision-making in today's data-driven world. However, the way data are visualized can influence and drastically change the conclusions people draw using data. The findings around visualization effectiveness are nuanced, and guidelines for effective visualization design depend on the visual channels used, chart types, and analysis tasks. This points to a significant need to understand the intersection of these factors to create optimized visualizations. We need a framework to define this intersection that fills the gap by providing a task-optimized visualization design for better quality and higher decision-making confidence that gives designers objective guidance. A task-optimized visualization design framework strategically integrates visual channels, visualization types, and specific low-level tasks to enhance data interpretation and optimize user task performance. We discuss constructing a visualization framework that considers both human perception for encoding techniques and the task being performed, enabling optimizing visualization design to maximize efficiency. Furthermore, we highlight a task-optimized framework's impact and potential applications.

Physical effort modulates perceptual awareness judgment independent of level of processing

Recent studies have emphasized the association between action and perceptual awareness, suggesting that action-related information can contribute to perceptual awareness. Given that the Level of Processing (LoP) hypothesis proposes that the emergence of awareness depends on the level of stimulus processing, the current study examines whether action impacts perceptual awareness across different processing levels. In Experiment 1, participants identified target stimuli’s color (low-level task) or category (high-level task) via mouse clicks, followed by visual awareness ratings. Experiment 2 replicated the tasks using hand-grip dynamometers. Results from Experiment 1 support the LoP theory, showing a more gradual emergence of awareness for low-level features and a more dichotomous emergence for high-level features. In Experiment 2, higher reported visual awareness ratings were observed at greater physical effort, regardless of task type. These results suggest that action-related information influences reported awareness of stimuli in the same way at low- and high-level stimulus processing.

Seeing the Trees: How a Concrete versus Abstract Mindset Improves Performance on Low-Level Assurance Tasks

ABSTRACT Prior research shows that individuals benefit from abstract construals when performing broader high-level assurance tasks. We extend this research by studying the effects of different construal mindsets on skepticism and performance in a low-level assurance task setting. Using a sample of 195 online participants, we find that individuals benefit from concrete (versus abstract) mindsets when performing low-level assurance tasks because concrete construals better allow them to focus on detailed evidence that may contradict client assertions. Specifically, we show that assurance providers who utilize a concrete (abstract) construal mindset when performing a low-level task display better (worse) skeptical judgments and actions. Taken together with findings from past research, our results support the notion that assurance providers should utilize the construal mindset that best matches the characteristics of the task at hand. Exploratory analysis suggests that client retention incentives may moderate the relationship between skeptical judgment and skeptical action in our context. Data Availability: Data are available from the authors upon request.

English Speakers’ Perception of Non-native Vowel Contrasts in Adverse Listening Conditions: A Discrimination Study on the German Front Rounded Vowels /y/ and /ø/

Previous research has shown that it is difficult for English speakers to distinguish the front rounded vowels /y/ and /ø/ from the back rounded vowels /u/ and /o/. In this study, we examine the effect of noise on this perceptual difficulty. In an Oddity Discrimination Task, English speakers without any knowledge of German were asked to discriminate between German-sounding pseudowords varying in the vowel both in quiet and in white noise at two signal-to-noise ratios (8 and 0 dB). In test trials, vowels of the same height were contrasted with each other, whereas a contrast with /a/ served as a control trial. Results revealed that a contrast with /a/ remained stable in every listening condition for both high and mid vowels. When contrasting vowels of the same height, however, there was a perceptual shift along the F2 dimension as the noise level increased. Although the /ø/-/o/ and particularly /y/-/u/ contrasts were the most difficult in quiet, accuracy on /i/-/y/ and /e/-/ø/ trials decreased immensely when the speech signal was masked. The German control group showed the same pattern, albeit less severe than the non-native group, suggesting that even in low-level tasks with pseudowords, there is a native advantage in speech perception in noise.

From silicon to solutions: AI's impending impact on research and discovery

The social sciences have long relied on comparative work as the foundation upon which we understand the complexities of human behavior and society. However, as we go deeper into the era of artificial intelligence (AI), it becomes imperative to move beyond mere comparison (e.g., how AI compares to humans across a range of tasks) to establish a visionary agenda for AI as collaborative partners in the pursuit of knowledge and scientific inquiry. This paper articulates an agenda that envisions AI models as the preeminent scientific collaborators. We advocate for the profound notion that our thinking should evolve to anticipate, and include, AI models as one of the most impactful tools in the social scientist's toolbox, offering assistance and collaboration with low-level tasks (e.g., analysis and interpretation of research findings) and high-level tasks (e.g., the discovery of new academic frontiers) alike. This transformation requires us to imagine AI's possible/probable roles in the research process. We defend the inevitable benefits of AI as knowledge generators and research collaborators—agents who facilitate the scientific journey, aiming to make complex human issues more tractable and comprehensible. We foresee AI tools acting as co-researchers, contributing to research proposals and driving breakthrough discoveries. Ethical considerations are paramount, encompassing democratizing access to AI tools, fostering interdisciplinary collaborations, ensuring transparency, fairness, and privacy in AI-driven research, and addressing limitations and biases in large language models. Embracing AI as collaborative partners will revolutionize the landscape of social sciences, enabling innovative, inclusive, and ethically sound research practices.

Metacognitive evaluation of postdecisional perceptual representations.

Perceptual confidence is thought to arise from metacognitive processes that evaluate the underlying perceptual decision evidence. We investigated whether metacognitive access to perceptual evidence is constrained by the hierarchical organization of visual cortex, where high-level representations tend to be more readily available for explicit scrutiny. We found that the ability of human observers to evaluate their confidence did depend on whether they performed a high-level or low-level task on the same stimuli, but was also affected by manipulations that occurred long after the perceptual decision. Confidence in low-level perceptual decisions degraded with more time between the decision and the response cue, especially when backward masking was present. Confidence in high-level tasks was immune to backward masking and benefitted from additional time. These results can be explained by a model assuming confidence heavily relies on postdecisional internal representations of visual stimuli that degrade over time, where high-level representations are more persistent.

CutFreq: Cut-and-Swap Frequency Components for Low-Level Vision Augmentation

Low-level vision plays a crucial role in a wide range of imaging quality and image recognition applications. However, the limited size, quality, and diversity of datasets often pose significant challenges for low-level tasks. Data augmentation is the most effective and practical way of sample expansion, but the commonly used augmentation methods in high-level tasks have limited improvement in the low-level due to the boundary effects or the non-realistic context information. In this paper, we propose the Cut-and-Swap Frequency Components (CutFreq) method for low-level vision, which aims to preserve high-level representations with directionality and improve image synthesis quality. Observing the significant frequency domain differences between reconstructed images and real ones, in CutFreq, we propose to transform the input and real images separately in the frequency domain, then define two stages for the model training process, and finally swap the specified frequency bands respectively and inversely transform to generate augmented samples. The experimental results show the superior performance of CutFreq on five low-level vision tasks. Moreover, we demonstrate the effectiveness of CutFreq in the low-data regime. Code is available at https://github.com/DreamerCCC/CutFreq.

Evaluation of the quality and the productivity of neuroradiological reading of multiple sclerosis follow-up MRI scans using an intelligent automation software

PurposeThe assessment of multiple sclerosis (MS) lesions on follow-up magnetic resonance imaging (MRI) is tedious, time-consuming, and error-prone. Automation of low-level tasks could enhance the radiologist in this work. We evaluate the intelligent automation software Jazz in a blinded three centers study, for the assessment of new, slowly expanding, and contrast-enhancing MS lesions.MethodsIn three separate centers, 117 MS follow-up MRIs were blindly analyzed on fluid attenuated inversion recovery (FLAIR), pre- and post-gadolinium T1-weighted images using Jazz by 2 neuroradiologists in each center. The reading time was recorded. The ground truth was defined in a second reading by side-by-side comparison of both reports from Jazz and the standard clinical report. The number of described new, slowly expanding, and contrast-enhancing lesions described with Jazz was compared to the lesions described in the standard clinical report.ResultsA total of 96 new lesions from 41 patients and 162 slowly expanding lesions (SELs) from 61 patients were described in the ground truth reading. A significantly larger number of new lesions were described using Jazz compared to the standard clinical report (63 versus 24). No SELs were reported in the standard clinical report, while 95 SELs were reported on average using Jazz. A total of 4 new contrast-enhancing lesions were found in all reports. The reading with Jazz was very time efficient, taking on average 2min33s ± 1min0s per case. Overall inter-reader agreement for new lesions between the readers using Jazz was moderate for new lesions (Cohen kappa = 0.5) and slight for SELs (0.08).ConclusionThe quality and the productivity of neuroradiological reading of MS follow-up MRI scans can be significantly improved using the dedicated software Jazz.

High-low level task combination for object detection in foggy weather conditions

For the object detection task in foggy weather conditions, image dehazing network is often used as preprocessing method to get a clear input. However, there is not strictly a strong positive correlation between image dehazing task and object detection task. Moreover, the preprocessing module can increase the inference time of the whole model to a certain extent. To alleviate these problems, we propose a novel High-Low level task combination network (HLNet) based on multitask learning, which can learn both high-level and low-level tasks. Specially, instead of restoring the features to clear pixel-wise feature space like common image dehazing method, we opt to perform a restoration in feature level to mitigate the influence of the Batch Normalization (BN) layer of encoder on dehazing task. HLNet jointly learn dehazing task and detection task in an end-to-end fashion, which ensures that the weather-specific information in latent feature space is suppressed. Moreover, we applied the HLNet framework on three different object detection networks, including RetinaNnet, YOLOv3 and YOLOv5s network, and achieved improvements of 1.7 percent, 2.3 percent, and 1.2 percent in mAP respectively. The experimental results demonstrate the effectiveness and generalization ability of our proposed HLNet framework in real foggy scenarios.

Modular toolsets for integrating HPC clusters in experiment control systems

New particle/nuclear physics experiments require a massive amount of computing power that is only achieved by using high performance clusters directly connected to the data acquisition systems and integrated into the online systems of the experiments. However, integrating an HPC cluster into the online system of an experiment means: Managing and synchronizing thousands of processes that handle the huge throughput. In this work, modular components that can be used to build and integrate such a HPC cluster in the experiment control systems (ECS) will be introduced. The Online Device Control library (ODC) [1] in combination with the Dynamic Deployment System (DDS) [2, 3] and FairMQ [4] message queuing library offers a sustainable solution for integrating HPC cluster controls into an ECS. DDS as part of the ALFA framework [5] is a toolset that automates and significantly simplifies a dynamic deployment of user-defined processes and their dependencies on any resource management system (RMS) using a given process graph (topology). ODC, in this architecture, is the tool to control and communicate with a topology of FairMQ processes using DDS. ODC is designed to act as a broker between a high level experiment control system and a low level task management system e.g.: DDS. In this work the architecture of both DDS and ODC will be discussed, as well as the design decisions taken based on the experience gained of using these tools in production by the ALICE experiment at CERN to deploy and control thousands of processes (tasks) on the Event Processing Nodes cluster (EPN) during Run3 as a part of the ALICE O2 software ecosystem [6].

A Practical Contrastive Learning Framework for Single-Image Super-Resolution.

Contrastive learning has achieved remarkable success on various high-level tasks, but there are fewer contrastive learning-based methods proposed for low-level tasks. It is challenging to adopt vanilla contrastive learning technologies proposed for high-level visual tasks to low-level image restoration problems straightly. Because the acquired high-level global visual representations are insufficient for low-level tasks requiring rich texture and context information. In this article, we investigate the contrastive learning-based single-image super-resolution (SISR) from two perspectives: positive and negative sample construction and feature embedding. The existing methods take naive sample construction approaches (e.g., considering the low-quality input as a negative sample and the ground truth as a positive sample) and adopt a prior model (e.g., pretrained very deep convolutional networks proposed by visual geometry group (VGG) model) to obtain the feature embedding. To this end, we propose a practical contrastive learning framework for SISR (PCL-SR). We involve the generation of many informative positive and hard negative samples in frequency space. Instead of utilizing an additional pretrained network, we design a simple but effective embedding network inherited from the discriminator network, which is more task-friendly. Compared with the existing benchmark methods, we retrain them by our proposed PCL-SR framework and achieve superior performance. Extensive experiments have been conducted to show the effectiveness and technical contributions of our proposed PCL-SR thorough ablation studies. The code and resulting models will be released via https://github.com/Aitical/PCL-SISR.

Anisotropic Scale-Invariant Ellipse Detection.

Detecting ellipses poses a challenging low-level task indispensable to many image analysis applications. Existing ellipse detection methods commonly encounter two fundamental issues. First, inferior detection accuracy could be incurred on a small ellipse than that on a large one; this introduces the scale issue. Second, inferior detection accuracy could be yielded along the minor axis than along the major one of the same ellipse; this leads to the anisotropy issue. To address these issues simultaneously, a novel anisotropic scale-invariant (ASI) ellipse detection methodology is proposed. Our basic idea is to perform ellipse detection in a transformed image space referred to as the ellipse normalization (EN) space, in which the desired ellipse from the original image is 'normalized' to the unit circle. With the establishment of the EN-space, an analytical ellipse fitting scheme and a set of distance measures are developed. Theoretical justifications are then derived to prove that both our ellipse fitting scheme and distance measures are invariant to anisotropic scaling, and thus each ellipse can be detected with the same accuracy regardless of its size and ellipticity. By incorporating these components into two recent state-of-the-art algorithms, two ASI ellipse detectors are finally developed and exploited to verify the effectiveness of our proposed methodology.

GarmentCode: Programming Parametric Sewing Patterns

Garment modeling is an essential task of the global apparel industry and a core part of digital human modeling. Realistic representation of garments with valid sewing patterns is key to their accurate digital simulation and eventual fabrication. However, little-to-no computational tools provide support for bridging the gap between high-level construction goals and low-level editing of pattern geometry, e.g., combining or switching garment elements, semantic editing, or design exploration that maintains the validity of a sewing pattern. We suggest the first DSL for garment modeling - GarmentCode - that applies principles of object-oriented programming to garment construction and allows designing sewing patterns in a hierarchical, component-oriented manner. The programming-based paradigm naturally provides unique advantages of component abstraction, algorithmic manipulation, and free-form design parametrization. We additionally support the construction process by automating typical low-level tasks like placing a dart at a desired location. In our prototype garment configurator, users can manipulate meaningful design parameters and body measurements, while the construction of pattern geometry is handled by garment programs implemented with GarmentCode. Our configurator enables the free exploration of rich design spaces and the creation of garments using interchangeable, parameterized components. We showcase our approach by producing a variety of garment designs and retargeting them to different body shapes using our configurator. The library and garment configurator are available at https://github.com/maria-korosteleva/GarmentCode.

Protection system planning in distribution networks with microgrids using a bi-level multi-objective and multi-criteria optimization technique

Microgrids are able to improve several features of power systems, such as energy efficiencies, operating costs and environmental impacts. Nevertheless, microgrids’ protection must work congruently with power distribution protection to safely take all advantages. This research contributes to enable their protection by proposing a bi-level method to simultaneously solve the allocation and coordination problems, where the proposed scheme also includes local protections of distributed energy resources. The uncertainties associated with generation and loads are categorized by the k-means method, as well. The non-dominated sorting genetic algorithm II is employed in the upper-level task to solve the protection and control devices allocation problem with two opposing objectives. In the lower-level task, a genetic algorithm ensures their coordination. Protection devices include reclosers and fuses from the network, and directional relays for the point of common coupling of microgrids, while control devices consist of remote-controlled switches. In contrast to related works, local devices installed at the point of coupling of distributed generation units are considered as well, such as voltage-restrained overcurrent relays and frequency relays. The optimal solution for the decision-maker is achieved by utilizing the compromise programming technique. Results show the importance of solving the allocation and coordination problems simultaneously, achieving up to $25,000 cost savings compared to cases that solve these problems separately. The integrated strategy allows the network operator to select the optimum solution for the protective system and avoid corrective actions afterward. The results also show the viability of the islanding operation depending on the decision maker's criteria.

Automatic Scatterplot Design Optimization for Clustering Identification.

Scatterplots are among the most widely used visualization techniques. Compelling scatterplot visualizations improve understanding of data by leveraging visual perception to boost awareness when performing specific visual analytic tasks. Design choices in scatterplots, such as graphical encodings or data aspects, can directly impact decision-making quality for low-level tasks like clustering. Hence, constructing frameworks that consider both the perceptions of the visual encodings and the task being performed enables optimizing visualizations to maximize efficacy. In this article, we propose an automatic tool to optimize the design factors of scatterplots to reveal the most salient cluster structure. Our approach leverages the merge tree data structure to identify the clusters and optimize the choice of subsampling algorithm, sampling rate, marker size, and marker opacity used to generate a scatterplot image. We validate our approach with user and case studies that show it efficiently provides high-quality scatterplot designs from a large parameter space.