Final Layer Research Articles

Nonlinear Deep Learning Framework for Precise Sentiment Classification in Textual Data

The study presents the Nonlinear Deep Learning Framework for Precise Sentiment Classification (NPSC), focusing on improving sentiment analysis in text data. The model addresses the challenge of capturing complex and hidden sentiment patterns using a combination of BERT embeddings, Bidirectional LSTM (BiLSTM), Convolutional Neural Networks (CNN), and an attention mechanism. This setup helps in identifying subtle relationships within text, enabling more precise sentiment predictions. NPSC tackles common issues faced by existing models that struggle with detailed text analysis. Using the IMDB movie review dataset, the approach relies on BERT to convert words into dense vectors, BiLSTM for capturing context from both directions, and CNN to extract key local features. An attention layer highlights important words, refining the sentiment detection process. The integration of these components in a nonlinear fusion layer allows for enhanced feature interaction, leading to accurate classification through a final softmax layer. The findings show that NPSC performs better across metrics like accuracy, precision, recall, and F1-score. Ablation studies reveal the impact of each component, demonstrating their role in enhancing the model’s effectiveness. Error analysis shows how NPSC manages mixed sentiments and challenging text patterns. The study highlights NPSC’s potential as a reliable method for sentiment analysis, contributing to better handling of complex textual data.

A Neural Phillips Curve and a Deep Output Gap

Many problems plague empirical Phillips curves (PCs). Among them is the hurdle that the two key components, inflation expectations and the output gap, are both unobserved. Traditional remedies include proxying for the absentees or extracting them via assumptions-heavy filtering procedures. I propose an alternative route: a Hemisphere Neural Network (HNN) whose architecture yields a final layer where components can be interpreted as latent states within a Neural PC. First, HNN conducts the supervised estimation of nonlinearities that arise when translating a high-dimensional set of observed regressors into latent states. Second, forecasts are economically interpretable. Among other findings, the contribution of real activity to inflation appears understated in traditional PCs. In contrast, HNN captures the 2021 upswing in inflation and attributes it to a large positive output gap starting from late 2020. The unique path of HNN’s gap comes from dispensing with unemployment and GDP in favor of an amalgam of nonlinearly processed alternative tightness indicators.

The role of vertical grid resolution and turbulent diffusion uncertainty on chemical transport modeling

Chemical transport models (CTM) tend to perform poorly in simulating pollution processes under weak turbulent diffusion conditions. In this study, we address this issue from the perspectives of vertical grid resolution and vertical mixing schemes. Three vertical grid resolution configurations (L4, L12, L40) with the CHIMERE model are evaluated during a winter episode, which includes a heavy pollution episode (PE) in the Paris region. The results emphasize the significance of vertical grid resolution, particularly noticeable during nighttime, and consequently impacts CHIMERE simulations under nocturnal stable conditions. Consistent improvement in CTM modeling is observed with refined vertical resolutions and the first layer height based on a simple linear vertical diffusion scheme defined as the initial Kz diffusion (IKD) scheme. Compared to the other two configurations, the finest configuration (referred to as L4-IKD) demonstrates an average improvement in root mean square error of 23.26 % and 25.09 % on regular days (RD) and 62 % and 129 % during PE, respectively. However, simulations using the 1.5-order turbulence kinetic energy (TKE) based eddy diffusivity closure scheme, named the new eddy diffusion (NED), are more sensitive to the first layer height setup. Excessively fine first-layer heights can lead to inaccurate TKE calculations. Generally, models with low vertical grid resolution can reasonably predict air quality on RD or during light pollution events but struggle with heavy PEs. One straightforward enhancement strategy involves adding an extra fine first layer height in CTM simulations, resulting in an average 50.10 % improvement from L4-IKD to L12-IKD during PE. Another strategy is enhancing the model's vertical diffusion scheme, which improved the CTM modeling by 26.67 % compared with IKD during PE under identical vertical grid resolution.

Baltisaksa huumor ja pastorianekdoodid

Baltic German humour and anecdotes about clergymen Keywords: anecdotes, Baltic Germans, humour Baltic German anecdotes (Pratchen) are a relatively unexplored part of Estonian folk humour, much like Baltic German culture within Estonian culture as a whole. This study focuses on anecdotes about Lutheran clergymen, which offer material for comparison with the more thoroughly researched Estonian counterparts. These humorous tales provide valuable insights into the cultural history, mentality, language as well as everyday life of the Baltic German community. Older Estonian anecdotes about clergymen often emphasize the social and ethnic differences between Estonian peasants and the clergy. In contrast, Baltic German anecdotes provide an internal perspective, exploring the role of the clergy within the Baltic German community and their outreach beyond it. The anecdotes can offer explanations for certain aspects that remain obscure and are therefore ridiculed in Estonian anecdotes. Two key features of Baltic German humour are its oral nature and the tradition of competing in witticisms (pliggern), a “national sport” possibly rooted in student culture. Initially shared orally, these anecdotes often referenced known people and places. They were later written down to preserve the memory and cultural history of the community. Older anecdotes published between the world wars, became classics that kept appearing in various forms in later collections. World War II added a final layer, with humour providing a way to encapsulate and process epochal events. In Baltic German anecdotes, clergymen are depicted in everyday activities and relationships, often taking on the role of the trickster parson. Unlike in Estonian jokes, the dynamic between the nobility and the church features prominently. This often takes the form of a humorous tug of war between the clergyman and the noble patron of the church, who may have attended school together in the past. The parson’s attitude towards parishioners is usually benevolent but patronizing, with the parish clerk, sexton, and coachman often playing supporting roles. Another central theme is the clergyman as a political figurehead, reflecting the tensions caused by the official campaigns of Russification and conversion to Russian Orthodoxy.

Optimization of MBE-grown GaSb buffer on GaAs substrates for infrared detectors

The aim of this work was to improve the quality of the GaSb buffer layers on GaAs substrates using the molecular beam epitaxy (MBE) technology. The high quality of the GaSb buffer layers is one of the most important elements enabling the synthesis of good quality of type II superlattices (T2SL) structures for infrared applications. The main challenges in this regard are: compensation of the difference in lattice constants between GaAs and GaSb and obtaining the highest achievable surface quality of the final GaSb layer. In the literature, many authors describe different techniques to obtain the best quality of a GaSb buffer layer. In this work, we present the results of HRXRD, AFM, TOF-SIMS, SEM, and Nomarski optical microscope measurements obtained for 2 μm thick GaSb buffer layers. The GaSb layers are made according to different techniques and these results are compared with a GaSb buffer construction technique according to our own technology. During the processes, we also obtained an unintentional structure of one of the buffer layers, which allowed us to obtain very good results in terms of surface structure and crystallographic quality where FWHM in ωRC scan was equal to 138 arcsec and RMS 0.20 nm proving that there is still a lot of work to be done in this area.

Hierarchical Fusion of Infrared and Visible Images Based on Channel Attention Mechanism and Generative Adversarial Networks.

In order to solve the problem that existing visible and infrared image fusion methods rely only on the original local or global information representation, which has the problem of edge blurring and non-protrusion of salient targets, this paper proposes a layered fusion method based on channel attention mechanism and improved Generative Adversarial Network (HFCA_GAN). Firstly, the infrared image and visible image are decomposed into a base layer and fine layer, respectively, by a guiding filter. Secondly, the visible light base layer is fused with the infrared image base layer by histogram mapping enhancement to improve the contour effect. Thirdly, the improved GAN algorithm is used to fuse the infrared and visible image refinement layer, and the depth transferable module and guided fusion network are added to enrich the detailed information of the fused image. Finally, the multilayer convolutional fusion network with channel attention mechanism is used to correlate the local information of the layered fusion image, and the final fusion image containing contour gradient information and useful details is obtained. TNO and RoadSence datasets are selected for training and testing. The results show that the proposed algorithm retains the global structure features of multilayer images and has obvious advantages in fusion performance, model generalization and computational efficiency.

Accurate detection of dilated cardiomyopathy onset through machine learning predictions from ECG data

Abstract Background Dilated cardiomyopathy (DCM) is a primary heart muscle disease characterised by left-ventricular dilatation and reduced contractility, often associated with arrhythmia and conduction disease. DCM often has a genetic component, putting family members of diagnosed patients at risk for the disease. Stratification tools to provide personalized screening advice are currently lacking. Artificial intelligence (AI) based analysis of the electrocardiogram (ECG) offers a potential solution to this challenge. Purpose Our study aims to develop and validate a novel AI-ECG model capable of prediction and early detection of DCM. Methods An AI-ECG model was developed on a United States (US) based secondary care centre dataset including 337,463 ECGs from 50,932 patients within 4 months of diagnostic echocardiography and 265,576 ECGs from 51,086 patients prior to the first diagnostic echo. The data was split 50/10/40% for training, validation, and testing, with each subset containing 9-12% DCM positive labels. The model architecture was based on a convolutional neural network with residual blocks with a modified final layer for a discrete-time survival model. The AI-ECG score was linked to changes in the median ECG within the test set to provide explainability. External validation was performed on 68,885 diagnostic ECGs (0.08% DCM positive) from the UK biobank (UKB) and a Chinese general hospital dataset with 317,854 diagnostic ECGs (2.2% DCM positive) and 76,936 predictive ECGs (1.7% DCM positive). We performed common and rare variant association studies for the AI ECG scores on quality-controlled UKB data. Results The AI-ECG score for predicted future DCM was associated with diverse ECG features (Fig. 1). In the US cohort test set, the model achieved high diagnostic accuracy with an AUC of 0.9 (95% CI 0.89-0.90) overall, and 0.89 (0.85-0.92) in a subset excluding ischemic or valvular heart disease. External datasets showed AUCs of 0.86 (0.81-0.91) in the UKB and 0.97 (96-0.97) in the Chinese dataset. The predictive performance of the AI-ECG score, adjusted for age and sex, closely approximated echocardiographic traits (LVEF and LVEDD, Fig. 2) with a concordance index (CI) of 0.79 (0.76-0.83) and 0.81 (0.77-0.85), respectively, at 10 years of follow up (FU). Combining AI-ECG scores with echocardiographic traits enhanced the CI to 0.84 (0.8-0.87). Similar findings were observed in the Chinese dataset, with a CI of 0.87 (0.85-0.89) for the AI-ECG score and 0.9 (0.88-0.92) for echocardiographic traits. Genetic analyses in UKB identified associations between the AI-ECG score and rare TTN variants as well as 25 common variant loci. Conclusion Our explainable AI ECG model for future DCM prediction demonstrates robust performance across datasets, providing comparable predictive accuracy to echocardiographic traits up to 10 years pre-diagnosis. Genetic associations support the model's validity and provide new tools for DCM genetic discovery.

Silicon-based integrated passive device stack for III-V/Si monolithic 3D circuits operating on RF band

In this study, we demonstrated a silicon (Si)-based integrated passive device (IPD) stack to support III-V/Si monolithic 3D (M3D) ICs operating on the radio frequency (RF) band. The IPD stack was fabricated based on an 8-inch CMOS process line and integrated via M3D with an InGaAs HEMT layer. A process condition for a trap rich layer and a buried oxide layer in the IPD was established to simultaneously minimizing both the RF loss and wafer bowing. Through the process condition, the RF loss of the coplanar waveguides was −0.631 dB/mm at 30 GHz, lower than that of the CMOS foundry, and the wafer bowing of the stack was as low as −5.5 μm. The maximum quality factor of the inductors showed good values when compared to those of other CMOS foundry process-based inductors operating on the RF bands reported thus far. To obtain a compressive profile for the IPD stack, which is one of the most important requirements in advancing to wafer-to-wafer-level 3D bonding with the III-V active layer, a process method for the final IMD layer of the IPD was developed, resulting in a change from a tensile profile to a compressive profile for the IPD (corresponding wafer bowing value from −12.6 to + 10.7 μm).

KAN-Transformer Model for UltraShort-Term Wind Power Prediction Based on EWMA Data Processing

When using the Transformer model for wind power prediction, the presence of noise in wind power data and the model’s final layer relying solely on a simple linear output reduces the model’s ability to capture nonlinear relationships, leading to a decrease in prediction accuracy. To address these issues, this paper proposes an ultrashort-term wind power prediction model based on exponential weighted moving average (EWMA) data processing and Kolmogorov–Arnold Network (KAN)-Transformer. First, multiple variable features are smoothed using EWMA, which suppresses noise while preserving the original data trends. Then, the EWMA-processed data is input into the Encoder and Decoder modules of the Transformer model to extract features. The output from the Decoder layer is then passed through the KAN layer, built using a cubic B-spline function, to enhance the model’s ability to capture nonlinear relationships, thereby improving the prediction accuracy of the Transformer model for wind power. Finally, experimental analysis is conducted, and it shows that the proposed model achieves the highest prediction accuracy, with a mean absolute error of 4.38 MW, a root mean squared error of 7.37 MW, and a coefficient of determination of 98.73%.

Attenuative Unified U-Net with Conditional PatchMatch Algorithm for Edge Preserving Salient Object Detection

The rise of Deep Neural Networks (DNNs) has significantly boosted salient object detection in computer vision tasks. However, downsampling methods like striding and pooling often introduce bluish artifacts near edges, hindering detection accuracy. To address this, the "Attenuative Unified U-Net with Conditional PatchMatch Algorithm" is proposed. The method employs Gaussian smoothing for noise reduction and depth refinement in preprocessing. In existing methods, the Edge-preserving salient object detection struggles with cluttered backgrounds and connected objects, like a bird on a rock, making it difficult to accurately distinguish edges and salient regions due to shared boundaries. Hence, an Attenuative Unified Backpropagated Entropy UNet is proposed, which integrates an Attention Mechanism that enhance feature map spatial weight assignment, emphasizing features or regions that are judged most relevant for edge and salient region detection. Then, the Cascaded Laplace pyramid is incorporated into the Unified U-Net’s design for multi-scale information processing, capturing contextual details effectively. The Backpropagated Entropy Loss Function is then created to ensure accurate and confident fused image output. The depth map, edge map, and salient area map are combined in the U-Net’s final output layer to create an overall fused image. Next, to addresses depth discontinuity problem a noteworthy novel idea the Conditional PatchMatch Random Field Algorithm (CPMRF), which combines PatchMatch efficiency and CRFs’ contextual modeling. PatchMatch iteratively propagates matches from neighboring patches to find similar patches in the remaining portion of the image and Conditional Random Field (CRF) is the next step, which improves the PatchMatch results by taking into account the connections between the matched patches and their neighboring areas. Hence, the experimental outcomes of the proposed model effectively show the improved detection of edge in salient object detection with better accuracy, precision, recall, MaxF, and F1 score with minimized MAE.

U-Net Semantic Segmentation-Based Calorific Value Estimation of Straw Multifuels for Combined Heat and Power Generation Processes

This paper proposes a system for real-time estimation of the calorific value of mixed straw fuels based on an improved U-Net semantic segmentation model. This system aims to address the uncertainty in heat and power generation per unit time in combined heat and power generation (CHPG) systems caused by fluctuations in the calorific value of straw fuels. The system integrates an industrial camera, moisture detector, and quality sensors to capture images of the multi-fuel straw. It applies the improved U-Net segmentation network for semantic segmentation of the images, accurately calculating the proportion of each type of straw. The improved U-Net network introduces a self-attention mechanism in the skip connections of the final layer of the encoder, replacing traditional convolutions by depthwise separable convolutions, as well as replacing the traditional convolutional bottleneck layers with Transformer encoder. These changes ensure that the model achieves high segmentation accuracy and strong generalization capability while maintaining good real-time performance. The semantic segmentation results of the straw images are used to calculate the proportions of different types of straw and, combined with moisture content and quality data, the calorific value of the mixed fuel is estimated in real time based on the elemental composition of each straw type. Validation using images captured from an actual thermal power plant shows that, under the same conditions, the proposed model has only a 0.2% decrease in accuracy compared to the traditional U-Net segmentation network, while the number of parameters is significantly reduced by 74%, and inference speed is improved 23%.

Multimodal Fusion of EEG and Audio Spectrogram for Major Depressive Disorder Recognition Using Modified DenseNet121.

This study investigates the classification of Major Depressive Disorder (MDD) using electroencephalography (EEG) Short-Time Fourier-Transform (STFT) spectrograms and audio Mel-spectrogram data of 52 subjects. The objective is to develop a multimodal classification model that integrates audio and EEG data to accurately identify depressive tendencies. We utilized the Multimodal open dataset for Mental Disorder Analysis (MODMA) and trained a pre-trained Densenet121 model using transfer learning. Features from both the EEG and audio modalities were extracted and concatenated before being passed through the final classification layer. Additionally, an ablation study was conducted on both datasets separately. The proposed multimodal classification model demonstrated superior performance compared to existing methods, achieving an Accuracy of 97.53%, Precision of 98.20%, F1 Score of 97.76%, and Recall of 97.32%. A confusion matrix was also used to evaluate the model's effectiveness. The paper presents a robust multimodal classification approach that outperforms state-of-the-art methods with potential application in clinical diagnostics for depression assessment.

A Multi-Layer Techno-Economic-Environmental Energy Management Optimization in Cooperative Multi-Microgrids with Demand Response Program and Uncertainties Consideration.

This paper presents a multi-layer, multi-objective (MLMO) optimization model for techno-economic-environmental energy management in cooperative multi-Microgrids (MMGs) that incorporates a Demand Response Program (DRP). The proposed MLMO approach simultaneously optimizes operating costs, MMG operator benefits, environmental emissions, and MMG dependency. This paper proposed a new hybrid ε-lexicography-weighted-sum that eliminates the need to normalize or scalarize objectives. The first layer of the model schedules MMG resources with DRP to minimize operating costs (local generation and power transactions with the utility grid) and maximize MMG profit. The second layer achieves the environmental operation of the MMG, while the third layer maximizes MMG reliability. This paper also proposed a new application of a recently developed enhanced equilibrium optimizer (EEO) for solving the three-layer EM problem. In addition, the uncertainties of solar power generation, wind power generation, load demand, and energy prices are considered based on the probabilistic 2m + 1 Point estimation method (PEM) approach. Three case studies are presented to verify the proposed MLMO approach on an MMG test system. In Case I, a deterministic EM is solved to simulate the MMG as a single layer to minimize costs and maximize benefits through DRP, while Case II solves the MLMO optimization problem. Simulation results show that the proposed MLMO technique reduces environmental emissions by 2.45% and 3.5% in its optimization layer and at the final layer, respectively. The independence index is also enhanced by 2.49% and 4.8% in its layer only and as a total increase, respectively. Case III is for the probabilistic EM simulation; due to the uncertain variables effect, the mean value in this case is increased by about 2.6% over Case I.

Effects of particle size, binder wettability, and sintering atmosphere on the microstructural and mechanical properties of binder jetted 18Ni300 alloy

A fully dense 18Ni300 mold steel was firstly prepared by the binder jetting (BJ) process. We developed a water-based binder characterized by its low viscosity, which significantly enhanced the penetration of binder into the fine powder layer. This process resulted in a deep, narrow cross-section due to reduced flow resistance, leading to strong inter-layer bonding and high surface quality of the green part. The densification behavior of 18Ni300 BJ green parts with different particle sizes was meticulously analyzed. A remarkable tensile property of YS = 783 MPa, UTS = 1012 MPa and EL = 8.8% were obtained in sintered 18Ni300 alloy with fine powder under 1460 °C and 50 mTorr vacuum. Building on thermodynamic theory, we introduced a novel criterion to assess the oxidation reactions in 18Ni300 parts fabricated by the BJ method during sintering under different atmospheric conditions.

Sub‐Milankovitch Signals in the Northern Polar Layered Deposits of Mars

AbstractOur understanding of the paleoclimate of Mars and the dominant forcing functions that drive large‐scale changes largely remains a mystery. However, the Martian Polar Layered Deposits (PLD) offer a promising avenue for unraveling the planet's recent paleoclimate history. Despite recent progress to detect a climate signal in the PLD, definitive evidence of a correlation between the stratigraphic record and Mars' recent orbital history remains elusive. We utilized new and updated techniques, including high‐resolution stratigraphic reconstruction from High Resolution Imagine Science Experiment stereo imagery and digital terrain models, combined with a technique of variable dip correction to account for three dimensional bedding orientations. Signal processing methods, such as wavelet and Fourier analysis, were applied to perform detailed time‐series analysis. These analyses revealed a quasi‐periodic signal indicative of fine bedding at a scale of approximately 2 m. Based on previously proposed deposition rates, these fine layers appear to correspond to timescales centering around 4,000 years. This suggests that the meter‐scale layers may not be the result of orbital forcing and hint at an unknown sub‐Milankovitch climate forcing mechanism. We discuss potential exogenic causal mechanisms, such as cyclic variations in solar activity, and endogenic factors, including large‐scale changes in dust distribution. Understanding the formation processes of these fine layers may significantly enhance our knowledge of Martian climate history and its driving forces.

Development of composite ultrafiltration membrane from fly ash microspheres and alumina nanofibers for efficient dye removal from aqueous solutions

In this work, a novel type of ultrafiltration ceramic membranes with the support based on fine fly ash microspheres and selective layer based on the alumina nanofibers with an aluminosilicate binder is proposed. The average pore sizes of the support and selective layer are 0.46 μm and 29 nm, respectively. The membrane is characterized by the compressive strength of 96 MPa and water permeability of 207 L m−2 h−1 bar−1. It is shown that the binder provides structural stability of selective layer and adhesion to the support. With increasing the binder content, the water permeability increases, reaches maximum, and then slightly decreases. The developed membranes are used for ultrafiltration of Blue Dextran dyes aqueous solutions with molecular weights of 70 kDa and 500 kDa and concentrations of 50 and 100 mg/L. The dyes rejection varies in the range 97–99 %, while the permeate flux is 100–140 L m−2 h−1 at the transmembrane pressure of 4 bars. The dye retention occurs via adsorption at the initial stage, which leads to the narrowing of pore size. Further, the dye filtration proceeds mainly due to size effects. The proposed membranes can be employed for dye removal from wastewater, and also allow chemical modification by carbon coating to be employed in electrochemically assisted ultrafiltration. The developed methodology promotes the recycling of thermal energy waste and introduces novel approaches to combine waste and synthesized raw materials in the production of low-cost ceramic membranes.

Relationship between Fine Layering of Stratified Water Environment and Vertical Turbulent Mass Transport

Relationship between Fine Layering of Stratified Water Environment and Vertical Turbulent Mass Transport

Study on the working performance and microscopic mechanism of high-performance grouting material (HPGM) reinforced dense fine sand layer

To address the challenges of small diffusion radius, low early strength and long setting time of traditional cement grout in dense fine sand layer, based on the synergistic mechanism, a novel high-performance grouting material (HPGM) was developed using ground granulated blast-furnace slag (GGBS), ultra-fine fly ash (UFFA), and ultra-fine Portland cement (UFPC) as raw materials. Firstly, Laboratory experiments were conducted to investigate the working performance of HPGM under varying proportions. Secondly, based on the self-designed advanced small pipe grouting full-scale test device, the diffusion performance and reinforcement characteristics of HPGM under varying proportions were analyzed. Finally, the XRD and SEM testing methods were employed to elucidated the hydration mechanism of the HPGM. The test results indicate that as the mass ratio of GGBS to UFFA decreases, there is a gradual increase in hardening rate, fluidity and setting time of HPGM. Additionally, the unconfined compressive strength (UCS) initially increase before decreasing. Furthermore, with the decreased mass ratio of GGBS to UFFA, the HPGM diffused in dense fine sand layer through compaction-splitting, permeation-splitting, and splitting grouting processes. The hydration products of the HPGM primarily consist of C(-A)-S-H, AFt, and Ca(OH)2, when the mass ratio of GGBS to UFFA falls within the range of 4:2–3:3, the UFFA facilitates the AFt hydration products in significant quantities. Its cementation and filling effect lead to the formation of a dense microstructure in the soil, resulting in a substantial reduction in permeability coefficient and a marked increase in UCS. The research findings can offer high-performance grouting material for underground engineering construction while also presenting innovative ideas for resource utilization from solid wastes like GGBS and UFFA.

Research Progress and Current Status of Gas-Solid Two-Phase Flow Technology in the Direction of Laser Cladding.

In the process of laser cladding, there are usually problems such as powder plugging and uneven delivery, which affect the quality of the final cladding layer. Therefore, powder convergence characteristics in laser cladding need to be further improved. Gas-solid two-phase flow technology has been widely used in the study of powder flow characteristics because it can precisely regulate the interaction between carrier air and powder flow. In this paper, we systematically review the current status of gas-solid two-phase flow in the field of laser cladding powder, deeply analyze the latest optimization progress of laser cladding nozzle design, and comprehensively explain the key progress of gas-solid two-phase flow technology in improving the uniformity and efficiency of powder field distribution. At the end of this paper, the research results are summarized and a series of prospective prospects are proposed, aiming to provide a valuable reference framework and directional guidance for the subsequent related research.

Visual search and real-image similarity: An empirical assessment through the lens of deep learning.

The ability to predict how efficiently a person finds an object in the environment is a crucial goal of attention research. Central to this issue are the similarity principles initially proposed by Duncan and Humphreys, which outline how the similarity between target and distractor objects (TD) and between distractor objects themselves (DD) affect search efficiency. However, the search principles lack direct quantitative support from an ecological perspective, being a summary approximation of a wide range of lab-based results poorly generalisable to real-world scenarios. This study exploits deep convolutional neural networks to predict human search efficiency from computational estimates of similarity between objects populating, potentially, any visual scene. Our results provide ecological evidence supporting the similarity principles: search performance continuously varies across tasks and conditions and improves with decreasing TD similarity and increasing DD similarity. Furthermore, our results reveal a crucial dissociation: TD and DD similarities mainly operate at two distinct layers of the network: DD similarity at the intermediate layers of coarse object features and TD similarity at the final layers of complex features used for classification. This suggests that these different similarities exert their major effects at two distinct perceptual levels and demonstrates our methodology's potential to offer insights into the depth of visual processing on which the search relies. By combining computational techniques with visual search principles, this approach aligns with modern trends in other research areas and fulfils longstanding demands for more ecologically valid research in the field of visual search.