17315 research outputs found
Sort by
Exploring visual search: past and current insights
No full textThe visual search paradigm is a fundamental concept in cognitive psychology and neuroscience that seeks to understand how people perceive and recognize specific items in visually complex environments. It also serves as a fundamental tool for understanding cognitive mechanisms such as attentional processes, information processing, and environmental awareness. This paradigm is widely used in laboratory settings with consistent stimuli to gain a better understanding of the factors that influence visual perception and attention. In addition, visual search plays a significant role in our daily lives, particularly in social interactions. Although the phrase visual search has been used in scientific literature since the 1940s, its use as a technique in visual cognitive science research only began in the 1970s. It was not until the 1980s, pioneered by Tresiman and Galade, that visual search became a research field on its own, gaining significant popularity among visual perception scientists in the early 1990s. The current definition of visual search, which remains accepted and quoted to date, was introduced by Wolfe and Horowitz in 2008. Since then, studies in the field have continued to be carried out without slowing down. Today, the visual search paradigm not only remains an active research area of cognitive psychology and neuroscience but also a diagnostic approach in neurodevelopmental disorders such as autism spectrum disorder and attention-deficit / hyperactivity disorder. In this paper, we review the evolution of the visual search paradigm from the 1980s to the present, discussing the principles and mechanisms that have been accepted by the scientific community thus far. Additionally, we review studies that have been conducted using this paradigm on the topics of visual perception and attention, categorizing them into research areas
Automatic detection of basic level categories [Temel seviye kategorilerin otomatik tespiti]
No full textBasic level categories (BLCs), which can be defined as the most inclusive level at which a concrete mental image of the entire category can be formed, have proven to be useful in a variety of applications in natural language processing (NLP) and computer vision (CV) tasks, such as word sense disambiguation, image searches, image description, and retrieval. Limiting their practical applications, current methods for detecting BLCs predominantly rely on rule-based methods and external knowledge sources rather than the information extracted directly from the text. In this manuscript, we propose a novel approach to detect BLCs that is merely based on the information obtained from the word embeddings, including Gaussian word embeddings (W2G) and embeddings retrieved from transformer-based models such as BERT and GPT-2. The proposed method significantly outperforms existing works in performance and practicality, demonstrating the effectiveness of contextual word embeddings for BLC detection
Multilabel contrastive learning based remote sensing scene classification via cosine similarity
No full textMulti-Label Classification is a fundamental task in remote sensing, which reflects real-world scenarios by enabling a sample to have more than one label. In the context of multi-class image classification, most recent methods usually make use of the contrastive learning strategy to increase the representative power of the backbone network. Yet, there has been little focus on generalizing the supervised contrastive learning strategy to multi-label classification tasks. In this paper, a new supervised contrastive loss function with a continuous modeling of inter-sample relations is proposed that outperforms common alternative strategies with an optical remote sensing scene classification dataset
Speed breeding of soybean by using 22 h photoperiod increases photochemical efficiency of pods and produces six generations per year
Full text linkFast generation cycling of plants has the potential to overcome the bottleneck of traditional breeding programmes, which often require several years to achieve the desired outcomes. Recent speed breeding methodologies have reduced generation times in both short- and long-day species by optimizing environmental conditions. However, protocols for short-day plants impose a constant short-day photoperiod throughout the entire life cycle, even though plants could benefit from extended light exposure. Here, we report a speed breeding scheme for soybean (Glycine max) based on a long-day photoperiod of 22 h (LD-22 h) applied upon flowering initiation (stage R1) using light-emitting diodes (LEDs) with a cool white (6000 K) and red light (660 nm) spectrum at 550 μmol/(m2s) photosynthetic photon flux at canopy level. We also outline an immature seed germination technique for early harvested green seeds collected from speed-breeding plants that markedly increased the germination rate. Combining these methods allowed our soybean speed breeding system to acquire a 92% germination rate from 58-day-old seeds, enabling six generations y−1 compared to typically only 1–3 using standard approaches. The impact of long photoperiods on soybean leaf and pod photochemical efficiency was examined. Although photosynthetic capacity (Vcmax, Jmax, and Amax) was significantly lower in leaves grown under LD-22 h photoperiod, seed production was unaffected, while PSII operating efficiency (Fq′/Fm′) in pods was markedly higher under LD-22 h compared to the SD-10 h photoperiod. Implementing our post-flowering long photoperiod conditions followed by an enhanced germination technique could facilitate rapid breeding for soybeans and be adapted for use with other photoperiod-sensitive short-day crops
Dynamic pricing when consumers have real options
No full textWe study optimal dynamic pricing under uncertainty in a platform ecosystem subject to technological uncertainty. We highlight that users joining the platform before the full development of the complementary goods and services obtain real options to benefit from future improvements in platform quality and network effects. The platform owner influences the network effects and equilibrium outcomes through its dynamic price policy that trades off building an earlier consumer base versus extracting rents from early adopters. A price-skimming policy is optimal when the cost of developing a complementary good is small. Interestingly, price-skimming becomes optimal when the development cost is high, as long as the value of the improved platform is either small or relatively high. For intermediate values, however, the platform adopts a price-penetration policy. Our paper provides new insights for building markets subject to the network effect under uncertainty
Mitigating information leakage in large language models: evaluating the impact of code obfuscation on vulnerability detection
No full textLarge Language Models (LLMs) have become widely used in software development, offering assistance in developing, debugging, and optimizing code. However, their ability to analyze code also raises security concerns, particularly regarding information leakage and vulnerability detection. When used for vulnerability analysis, LLMs may inadvertently expose sensitive security weaknesses, effectively leaking information that could be exploited by attackers. This study investigates how effectively LLMs detect security vulnerabilities in source code and assesses the impact of various obfuscation techniques on reducing this risk. We constructed a data set of 400 C and Python code snippets, each containing security vulnerabilities classified into 51 Common Weakness Enumeration (CWE) categories. These snippets were analyzed using the ChatGPT-4o mini API to measure vulnerability detection accuracy before and after applying obfuscation techniques, including comment removal, string manipulation, identifier renaming, control flow and data flow obfuscation, dead code insertion, full encoding, and LLM-generated obfuscation. Our results demonstrate that while LLMs can effectively reason about the code, obfuscation significantly reduces their detection capabilities of potential vulnerabilities in the code, which we aim to prevent from leaking. Among the techniques tested, dead code insertion and control flow obfuscation were the most effective in decreasing detection accuracy, whereas simpler methods such as comment and identifier obfuscation had minimal impact. Additionally, encoding-based obfuscation, though highly disruptive, proved impractical due to severe functionality loss. These findings emphasize the need to balance obfuscation for security with maintaining code usability. By evaluating the effectiveness of different obfuscation techniques, this research provides practical guidance for developers seeking to minimize information leakage while leveraging LLMs in software development
Sectoral heterogeneity in wage stickiness and monetary policy transmission
No full textIt is well known that introducing sectoral heterogeneity in price stickiness amplifies monetary non-neutrality in standard New Keynesian models (Carvalho, 2006). Yet, less attention has been paid to the role of heterogeneity in wage stickiness. Using industry-level data, I document a statistically significant negative correlation of –0.27 between wage and price rigidity at the 3-digit NAICS level. I then develop a multi-sector New Keynesian model that incorporates heterogeneity in both wage and price rigidities. The model shows that, depending on the correlation between sectoral wage and price rigidities, the cumulative real response to a monetary shock can either be amplified or dampened relative to a benchmark economy with only heterogeneous price rigidity. Specifically, a perfectly positive correlation between sectoral wage and price rigidities amplifies the cumulative real response by up to 14 percent, whereas a perfect negative correlation reduces it by approximately 9 percent. However, in the model empirically calibrated to the 53-sector U.S. economy, the aggregate impact of wage rigidity heterogeneity is limited, as the weak observed correlation and the influence of large, moderately rigid sectors mute the underlying transmission channel
Mütareke dönemine sol edebiyattan bir bakış: kan konuşmaz örneği
No full textIn most novels dealing with the Occupation of Istanbul, Ankara—the headquarters of the National Struggle—represents dignity, Turkishness, heroism, and the nascent nation-state, while occupied Istanbul embodies falsehood, betrayal, the collapsing Ottoman Empire, hedonism, cosmopolitanism, and non-Turkish elements. In other words, the vast majority of works in this corpus present occupied Istanbul as the antithesis of National Struggle Ankara, thereby reinforcing the sanctity and grandeur of the Ankara-centered resistance narrative.
However, this framework cannot be applied to all works addressing the Armistice period; some offer alternative perspectives on the era and its dynamics. For instance, Nazım Hikmet’s novel Kan Konuşmaz (Blood Does Not Speak), the focus of this article, examines this decade of war—and particularly Armistice-era Istanbul—from a socialist perspective. This article will analyze the novel’s approach and its contrasts with nationalist narratives
Impact of asymmetricity of indexable cutter bodies on chatter resistance
Full text linkMilling operations are often limited by regenerative chatter vibrations arising due to the flexible machine tool parts which have dynamic effects on both stationary and rotary directions. Additionally, asymmetricity (or, non-axisymmetricity) of the milling cutter has an impact on rotating dynamics of the machine tool. However, asymmetric cutters may not always have positive effects; hence, care must be taken when using them. This paper investigates how to achieve desired chatter stability characteristics by tuning the dynamic response of the tool body. Both frequency and time domain methods were used to predict the stability lobes diagram according to the measured tool tip frequency response functions. An indexable end mill with exchangeable head (namely, tool bit) is used for demonstration of dynamic tuning. Additionally, receptance coupling substructure analysis is used to predict the tool tip frequency response function for various degrees of asymmetry. The down milling cutting tests validated that the proposed approach captures the sensitivity of the stability lobes diagram to the degree of asymmetricity
From insulator to metal: theoretical assessment on the optical properties of vanadium dioxide using many-body first-principles approaches
No full textVanadium dioxide (VO2) exhibits a temperature-driven insulator-to-metal transition, making it a promising material for optical and electronic applications. In this study, we perform a systematic first-principles investigation of the electronic and optical properties of VO2in its monoclinic (M1) and rutile (R) phases using density functional theory (DFT), many-body perturbation theory (G0W0), and the Bethe-Salpeter equation (BSE). Our results reveal that excitonic effects play a crucial role in accurately describing the dielectric response of the semiconducting M1phase, with G0W0/BSE and PBE/BSE approaches yielding optical spectra in excellent agreement with experimental data. For the metallic R phase, we find that the random phase approximation (RPA) at the PBE level provides a reliable description of its optical properties, particularly in the visible range, as long as intraband contributions are included. The frequency-dependent dielectric functions presented in this work achieve the required accuracy for large-scale optical simulations relevant to smart coatings and tunable infrared devices. To support further research and applications, we provide our computed optical data in an open-access repository on ZENODO