.png?v=trim20261018)
1CONN Creative Reklam Ajansı
Sektörünüzle ilgili genel ve yerel bilgileri büyük bir titizlikle araştırıyor, sizin için gerçekleştireceğimiz çalışmalarımıza bu bilgiler ışığında yön veriyoruz. İlk düşündüğümüz; “Biz, bu markaya ne katabiliriz?”
LER-YOLO: Reliability-Aware Expert Routing for Misaligned RGB-Infrared UAV Detection
Detecting small unmanned aerial vehicles from RGB-infrared remote-sensing pairs remains challenging due to tiny target scale, cluttered backgrounds, and spatial misalignment between heterogeneous sensors. Existing bimodal detectors often align or fuse features without assessing the reliability of local cross-sensor correspondence, allowing mismatch artifacts to propagate into the detection head. To address this issue, we propose LER-YOLO, a reliability-aware sparse mixture-of-experts framework for misaligned RGB-infrared UAV detection. LER-YOLO first introduces an Uncertainty-Aware Target Alignment module that resamples visible features toward the infrared reference and estimates a spatial reliability map. This reliability prior is then used by a Reliability-Guided Sparse MoE Fusion module to adaptively select k experts from RGB-dominant, infrared-dominant, and interactive fusion experts, enabling trustworthy cross-modal interaction while suppressing unreliable fusion. Experiments on the public MBU benchmark under a YOLOv5s-family protocol show that LER-YOLO achieves 89.7+/-0.2% AP50 over three independent seeds, with a best result of 89.9%. Extensive ablations, parameter-matched …
Dévissages des F-complexes de D-modules arithmétiques en F-isocristaux surconvergents
Nous définissons et étudions les dévissages des $F$-complexes de $\mathcal{D}$-modules arithmétiques en $F$-isocristaux surconvergents. Nous prouvons que les $F$-complexes surholonomes sont dévissables en $F$-isocristaux surconvergents. On établit ensuite une formule cohomologique, étendant celle d'Étesse et Le Stum, des fonctions $L$ associées aux duaux des $F$-complexes de $\mathcal{D}$-modules arithmétiques dévissables en $F$-isocristaux surconvergents. Puis, nous obtenons un analogue $p$-adique de Weil II étendant celui de Kedlaya.
Une représentation en graphe pour l'enseignement de XML
Currently, XML is a format widely used. In the context of computer science teaching, it is necessary to introduce students to this format and, especially, at its eco-system. We have developed a model to support the teaching of XML. We propose to represent an XML schema as a graph highlighting the structural characteristics of the valide documents. We present in this report different graphic elements of the model and the improvements it brings to data modeling in XML.---XML est un format actuellement très utilisé. Dans le cadre des formations en informatique, il est indispensable d'initier les étudiants à ce format et, surtout, à tout son éco-système. Nous avons donc mis au point un modèle permettant d'appuyer l'enseignement de XML. Ce modèle propose de représenter un schéma XML sous la forme d'un graphe mettant en valeur les caractéristiques structurelles des documents valides. Nous présentons dans ce rapport les différents éléments graphique du modèle et les améliorations qu'il apporte à la modélisation de données en XML.
Sociophysics in the mexican case (Sociofísica en el caso mexicano)
The availability of electoral data in electronic format allowed to physicist and mathematicians their own analysis. Beyond theoretical models for electoral processes, new regularities had been found in the data analysis of many countries and many elections. Here, we discuss some of the statistical regularities found for México, from power laws Daisy models linked in corporate vote. The methods used are those from standard statistics and statistical mechanics. As a particular application we discuss an entropic approach to the distribution of votes for the Partido Revolucionario Institucional and we demonstrate that this kind of method DOES NOT allow to discover a massive buy of vote. The argument is that such a distribution has been amazingly consistent in the last two decades elections. Other methods have been unsuccessful to discover signal of fraud. RESUMEN: Con la disponibilidad de datos electorales en formato electrónico los físicos y matemáticos iniciaron sus propios análisis de los datos. Más allá de los modelos teóricos desarrollados para los procesos electorales, nuevas regularidades empezaron a surgir de los datos en muchas elecciones y en muchos países. Aquí discutimos …
ATM: Why Latent World Models Can Fail to Plan
Latent world models can achieve accurate latent prediction yet still differ substantially in downstream planning performance. We argue that a key source of this discrepancy lies in the structure of action-induced latent transitions. We formalize action-identifiability through Bayes inverse risk, characterizing how much uncertainty about an action remains after observing the transition it induces. Model-predicted transitions can become highly self-decodable while encoding a domain-specific action relationship that fails to transfer to real environment transitions. We characterize this mismatch through cross-domain inverse transfer, instantiated as the Action-Consistency Transfer Matrix (ATM), a $2\times2$ inverse-risk matrix over real and predicted transition domains. Across TwoRoom, PushT, and OGBench-Cube, true-transition action-identifiability tracks downstream planning substantially better than standard prediction loss, while the full ATM reveals highly self-decodable yet cross-domain-inconsistent predicted transitions. Controlled interventions on the two domains further produce distinct transition structures and planning outcomes, supporting this decomposition. The same …
An Eco-Friendly Tourism Trip Planner Driven by Z-Number Uncertainty: Evidence from Baku, Dubai, Abu Dhabi, Tashkent, and Istanbul
Deployed tourism trip planners ingest travel times and ratings as exact numbers, although routing services return travel times that vary with the hour of departure and a platform rating may rest on a handful of reviews or on hundreds of thousands. We present an eco-friendly trip planner that represents both quantities as Z-numbers, which pair a fuzzy restriction on a value with a fuzzy reliability on the evidence behind it, and that constructs them directly from routing-service observations at six departure hours and from platform review counts rather than from linguistic elicitation. A hybrid loss combines score, carbon, travel-time, and reliability terms, and an adaptive large neighborhood search returns single-day itineraries through a conversational front end. We compare the planner with a crisp-mean baseline under an identical optimizer in Baku, Dubai, Abu Dhabi, Tashkent, and Istanbul, with fifteen points of interest and thirty seeds per city, and we evaluate the selected itineraries on their trip carbon and their total travel time. Pooled over 150 seed-paired runs, the planner lowers trip carbon from 3.24 to 2.61 kg (19.3%, Wilcoxon p = 0.017; city-clustered bootstrap …
VR with Older Adults: Participatory Design of a Virtual ATM Training Simulation
In this paper we report on a study conducted with a group of older adults in which they engaged in participatory design workshops to create a VR ATM training simulation. Based on observation, recordings and the developed VR application we present the results of the workshops and offer considerations and recommendations for organizing opportunities for end users, in this case older adults, to directly engage in co-creation of cutting-edge ICT solutions. These include co-designing interfaces and interaction schemes for emerging technologies like VR and AR. We discuss such aspects as user engagement and hardware and software tools suitable for participatory prototyping of VR applications. Finally, we present ideas for further research in the area of VR participatory prototyping with users of various proficiency levels, taking steps towards developing a unified framework for co-design in AR and VR.
Catégories de foncteurs en grassmanniennes
Soit F la catégorie des foncteurs entre espaces vectoriels sur un corps fini. Les catégories de foncteurs en grassmanniennes sont obtenues en remplaçant la source de cette catégorie par la catégorie des couples formés d'un espace vectoriel et d'un sous-espace. Ces catégories possèdent une très riche structure algébrique ; nous étudierons notamment leurs objets finis et leurs propriétés cohomologiques. Nous donnons des applications à la filtration de Krull de la catégorie F et à la K-théorie stable des corps finis. -- Let F be the category of functors between vector spaces over a finite field. The grassmannian functor categories are obtained by replacing the source of this category by the category of pairs formed by a vector space and a subspace. These categories have a very rich algebraic structure; we study in particular their finite objects and their homological properties. We give applications to the Krull filtration of the category F and to the stable K-theory of finite fields.
ATM: CID-Brokered Pre-Write Admission for Multi-Agent Code Co-Synthesis
Multi-agent LLM systems can decompose software-engineering work into planning, generation, validation, and repair, but a narrower systems problem remains: before any governed shared mutation is applied, a system must decide which concurrently formed write intents may proceed in parallel, which require deterministic composition or serialization, and which must take a fail-closed path. We address this problem with the AI-Atomic-Framework (ATM), a specification-grounded governance substrate for software agents operating within a single governance domain. ATM binds task intent, repository scope, write admission, validation, and evidence obligations into one governance chain. A Content Identifier (CID) broker serves as the shared-mutation admission subsystem. Adapter-guided atomization maps write intents to semantic atoms and bounded regions; when persistent atom-map coverage is incomplete, virtual atoms provide temporary auditable governance units for conservative comparison and routing. Governed shared writes are ultimately applied by a neutral steward rather than directly by proposing agents. Evaluation combines controlled, field, adoption, and extension evidence, including a …
Evaluating LLMs' Effectiveness on Real-World Consumer Device Repair Questions
Consumer device repair is an important but underexplored testbed for large language models (LLMs). Repair tasks require reasoning over incomplete problem descriptions, hardware-specific diagnostics, actionable troubleshooting, and safety-critical decisions, where incorrect advice can cause device damage, battery hazards, or permanent data loss. We introduce a benchmark of 991 real-world repair questions from Reddit spanning phone repair, computer repair, and data recovery, each paired with technician-written reference solutions, and provide Bangla translations to evaluate cross-lingual performance. We evaluate six state-of-the-art LLMs in English and Bangla using four repair-specific criteria: correctness, completeness, practicality, and safety. Our results show that while LLMs can provide useful repair assistance, they remain unreliable for high-risk real-world repair tasks without rigorous evaluation and explicit safety safeguards. Phone repair is the most difficult and safety-sensitive domain, and all models make substantial errors in board-level diagnosis, repair prioritization, and safe recovery procedures. Across domains and models, Bangla responses consistently perform …
1Klinik CMS Web Yönetim Sistemi
Yeni Nesil Klinik Deneyimi. Bilim, teknoloji ve estetiği modern hasta deneyimiyle bir araya getiriyoruz. Dijital planlama, kişiye özel tedavi yaklaşımı ve güçlü hasta iletişimi.
Dentastik Klinik Yönetimi
Kliniğinizin tüm dünyası Tek bir akış. İlk temastan son kontrole. Satıştan tedaviye. Birbirine bağlı her an, tek ekosistemde.
