5916 research outputs found
Sort by
NTIRE 2023 challenge on night photography rendering
No full textThis paper presents a review of the NTIRE 2023 challenge on night photography rendering. The goal of the challenge was to find solutions that process raw camera images taken in nighttime conditions conditions, and thereby produce a photo-quality output images in the standard RGB (sRGB) space. Unlike the previous year's competition, participants were not provided with a large training dataset for the target sensor. Instead, this time they were given images of a color checker illuminated by a known light source. To evaluate the results, a sufficient number of viewers were asked to assess the visual quality of the proposed solutions, considering the subjective nature of the task. The highest ranking solutions were further ranked by Richard Collins, a renowned photographer. The top ranking participants' solutions effectively represent the state-of-the-art in nighttime photography rendering
Between solidarity and conflict: Tactical biosociality of Turkish egg donors
No full textBased on ethnographic fieldwork and interviews conducted with Turkish egg donors at a Northern Cypriot clinic, this article investigates tactical biosociality of cross-border egg donors that allows them to manage social relations and orient themselves in transnational egg donation (including the processes from recruitment to self-management in and beyond the clinic) under legally restrictive and socially stigmatizing conditions. Addressing the social and collective dimensions of tactics and recognizing the fragmented and conflictual forms of biosociality, it aims to shed light on the complex and ambivalent aspects of tactical biosociality in relation to selective disclosure and stigma within the context of transnational egg donation. Tactical biosociality involves possibilities for solidarity and alliances, and also for conflict and competition among egg donors. It is because for young Turkish women, egg donation retains both gendered moral and financial values that must be tactically negotiated while navigating the wider context of heteropatriarchal cultural norms and expectations, precarious economic and social conditions, biomedical profit and biopolitical control.National Science Foundatio
The effect of few historical data on the performance of sample average approximation method for operating room scheduling
No full textWe model the scheduling problem of a single operating room for outpatient surgery, with uncertain case durations and an objective function comprising waiting time, idle time, and overtime costs. This stochastic scheduling problem has been studied in diverse forms. One of the most common approaches used is the sample average approximation (SAA). Our contribution is to study the use of SAA to solve this problem under few historical data using families of log t distributions with varying degrees of freedom. We analyze the results of the SAA method in terms of optimality convergence, the effect of the number of scenarios, and average computational time. Given the case sequence, computational results demonstrate that SAA with an adequate number of scenarios performs close to the exact method. For example, we find that the optimality gap, in units of proportional weighted time, is relatively small when 500 scenarios are used: 99% of the instances have an optimality gap of less than 2.6 7% (1.74%, 1.23%) when there are 3 (9, many) historical samples. Increasing the number of SAA scenarios improves performance, but is not critical when the case sequence is given. However, choosing the number of SAA scenarios becomes critical when the same method is used to choose among sequencing heuristics when there are few historical data. For example, when there are only three (nine, many) historical samples, 99% of the instances have less than 25.38% (13.15%, 6.87%) penalty in using SAA with 500 scenarios to choose the best sequencing heuristic
Deterministik doğrusal eşik modelinde n-hop etki maksimizasyon problemi
No full textThe Influence Maximization Problem (IMP) finds a set of highly influential nodes within a social network in order to maximize the spread of influence. We consider that people can have influence on their direct (1-hop), 2-hop and 3-hop neighbors. IMP with extended influence transitivity is called n-hop IMP. In this paper, we study the problem under the deterministic linear threshold model and propose a heuristic solution. In our proposed heuristic model, there are two parts, extended seed set algorithm and local search. Main purpose of extended seed set algorithm is creating a node set and send it to local search which is based on trying to improve it via replacing the nodes in the given set. We used two different features for selecting the candidate nodes. We propose an equation to estimate the value of a node set without actually computing. We used real-life and synthetic networks to test our solution method and generated weight and threshold values in three different methods.Etki Enbüyükleme Problemi, bir sosyal ağ içinde etkinin yayılmasını en üst düzeye çıkarmak için bir dizi son derece etkili düğüm bulur. İnsanların doğrudan (1-adım), 2-adım ve 3-adım komşuları üzerinde etkileri olabileceğini düşünüyoruz. Genişletilmiş etki geçişliliğine sahip Etki Enbüyükleme Problemi, n-adım Etki Enbüyükleme Problemi olarak adlandırılır. Bu yazıda, bahsedilen problemi belirlenmiş doğrusal eşik modeli altında inceliyoruz ve sezgisel bir çözüm buluyoruz. Önerdiğimiz sezgisel modelimizde, genişletilmiş çekirdek küme algoritması ve yerel arama olmak üzere iki bölüm var. Genişletilmiş çekirdek küme algoritmasının temel amacı, bir düğüm kümesi oluşturmak ve onu verilen kümedeki düğümleri değiştirerek iyileştirmeye dayanan yerel aramaya göndermektir. Aday düğümleri seçmek için iki farklı özellik kullandık. Gerçekte hesaplama yapmadan bir düğüm kümesinin değerini tahmin etmek için bir denklem buluyoruz. Çözüm yöntemimizi test etmek için gerçek hayat ve sentetik ağları kullandık ve üç farklı yöntemde ağırlık ve eşik değerleri oluşturduk
Deep learning-based blind image super-resolution with iterative kernel reconstruction and noise estimation
No full textBlind single image super-resolution (SISR) is a challenging task in image processing due to the ill-posed nature of the inverse problem. Complex degradations present in real life images make it difficult to solve this problem using naïve deep learning approaches, where models are often trained on synthetically generated image pairs. Most of the effort so far has been focused on solving the inverse problem under some constraints, such as for a limited space of blur kernels and/or assuming noise-free input images. Yet, there is a gap in the literature to provide a well-generalized deep learning-based solution that performs well on images with unknown and highly complex degradations. In this paper, we propose IKR-Net (Iterative Kernel Reconstruction Network) for blind SISR. In the proposed approach, kernel and noise estimation and high-resolution image reconstruction are carried out iteratively using dedicated deep models. The iterative refinement provides significant improvement in both the reconstructed image and the estimated blur kernel even for noisy inputs. IKR-Net provides a generalized solution that can handle any type of blur and level of noise in the input low-resolution image. IKR-Net achieves state-of-the-art results in blind SISR, especially for noisy images with motion blur.TÜBİTA
NFT primary sale price and secondary sale prediction via deep learning
No full textNon Fungible Tokens (NFTs) are blockchain-based unique digital assets defining ownership deeds. They can characterize various different objects such as collectible, art, and in-game items. In general, NFTs are encoded by blockchains smart contracts, and they are traded via cryptocurrencies. Their price and investors attention on them has remarkably increased especially in 2021, making them promising alternative class of investment. Surprisingly, predicting their prices has only recently started to be analyzed systematically. In this paper, we focus on predicting NFT primary sale price and secondary sale via deep learning. We use multimodal data, NFT images and NFT text characteristics when predicting their prices. Here, we show that contrasting the different and similar (DS) hierarchical features of images and text serves as an important identifying marker for their price, with the consequence that we only need to direct our attention to this aspect when designing a multimodal NFT price predictor. When designing NFT price predictor from multimodal data without using any financial attributes, we come up with Fine-Grained Differences-Similarities Enhancement Network (FG-DSEN), which improves detection with a simple and interpretable structure to enhance the DS aspect between images and text. According to detailed assessment on publicly available NFT dataset, our proposed approach outperforms baselines on both price direction prediction and secondary sale participation prediction according to several machine learning classification metrics
Sensorless position control of solenoid actuators for soft landing using super-twisting sliding mode control
No full textThis article presents an open-loop control methodology to achieve lower seating velocities (i.e. soft-landing) for solenoid-based injector systems which are widely used in automotive as fuel injection valves or gas exchange valves of internal combustion engines to spray various fluids. Physical sensors are not preferred to be used in injectors in order to increase reliability and reduce cost. As a result, it becomes impossible to control the motion of the moving parts within injectors in closed-loop. This study offers a novel sensorless position tracking approach with which impact noise can be reduced and mechanical wear and tear can be minimized. Using the Hammerstein-Wiener modeling method and a super-twisting sliding mode controller this new approach replicates the dynamics of the injector and tracks specially designed position reference signals to achieve soft landing. The effectiveness of this approach is based on the observed negligible position and velocity errors between the estimated and actual measurements. This study also offers a new way to optimize the settling time of the injector systems, while ensuring soft landing. Using the proposed approach here, the closing profiles of the reference signals were refined according to the admittance time of the solenoid actuator and the optimal closing profile signals were selected based on performance comparisons with the baseline. The results of the experiments are presented and the promising effectiveness of the proposed approach is discussed.TÜBİTAKPre-prin
A multioctave 8 GHz-40 GHz receiver for radio astronomy
No full textAccurate measurement of angular positions on the sky requires a well-defined system of reference, something that in practice is realized by the International Celestial Reference Frame (ICRF) with observations of distant (typical redshift similar to 1) Active Galactic Nuclei (AGN). At such great distances a subset of these objects exhibit as little as 10-50 mu as/year observed parallax or proper motion, thus giving the frame excellent spatial and temporal stability. Until fairly recently the majority of AGN centered imaging was accomplished in the S (2.3 GHz) and X (8.4 GHz) radio frequency bands, however S-band observations for reasons such as sensitivity "plateauing", increased source structure (jets), and radio frequency interference (RFI) have become less productive. Following spacecraft telemetry moves to higher frequencies and a desire to strengthen JPL's leadership in defining the next-generation of celestial reference frames has motivated the development of a "Quad-band" prototype receiver that operates in X, Ku, K, and Ka band in both right hand (RCP) and left hand (LCP) circular polarization. The goal of this receiver is to achieve less than a 20 % increase in noise over the Jansky Very Large Array (JVLA, NRAO) performance specification, which in such a wide bandwidth represents a revolutionary capability. To evaluate the various technical developments of the 8 GHz-40 GHz receiver the feedhorn optical beam was designed to interface to the US based Very Long Baseline Array (VLBA). The receiver's intermediate frequency (IF) spans 4 GHz-8 GHz, giving rise to up to eight 4 GHz IF channels for a fully populated instrument. This paper outlines the technical development of a 21/2 octave wide (8 GHz-40 GHz) X-Ka band prototype receiver, fulfilling a need for super broadband technology within the VLBI network. An important additional benefit of the wideband receiver approach is its simplicity and low cost of operation.Jet Propulsion Laboratory, California Institute of Technology ; National Aeronautics & Space Administration (NASA)Publisher versio
Spurred or spurned? The effects of combined innovations and major obstacles on B2B SME capability generation and future focus
No full textExpanding organizational learning theory and ambidextrous theory, this research investigates the potential differences in connections between business-to-business (B2B) small to medium enterprises (SME) engaging in customer facing (product and service) innovations versus B2B SMEs engaging in process innovations in addition customer facing innovations related to organizational capability generation (innovation, business model) and future focus on innovations or on investing in facilities and machinery. It also examines to what extent the context of major obstacles in financing and/or competition impact the same elements for the companies. Using survey data from leaders at 1405 B2B SMEs in the U.K., the results show that some of the elements spur improvements in the capabilities and focus while other combinations spurn the level of capabilities and focus
Çizelgeleme ve yönlendirme problemlerinde büyük ölçekli optimizasyon yöntemlerinin uygulanması
No full textIn this thesis, we consider three different applications of large-scale optimization methods. We focus on the blood donation tailoring problem under uncertain demand in the first problem. In the second one, we propose a model for hybrid manufacturing consisting of flexible manufacturing systems and typical manufacturing machines. In the last one, we consider a two-echelon vehicle routing problem for last-mile delivery of groceries. In the first part of the thesis, we propose a stochastic scenario-based reformulation of the blood donation management problem that adopts multicomponent apheresis and utilizes donor pool segmentation as here-and-now and wait-and-see donors. The donation pool segmentation enables more flexible donation schedules than the orthodox donation approach because wait-and-see donors may adjust their donation schedules according to the realized values of demand over time. We propose a column generation approach to solve the associated multi-stage stochastic donation tailoring problem for realistically sized instances. The second part considers a flexible/hybrid manufacturing production setting with typically dedicated machinery to satisfy regular demand and a flexible manufacturing system to handle surged demand. We model the uncertainty in demand using a scenario-based approach and allow the business to make here-and-now and wait-and-see decisions exploiting the cost-effectiveness of the standard production and responsiveness of the flexible manufacturing systems. We propose a branch-and-price algorithm as the solution approach. Our computational analysis shows that this hybrid production setting provides highly robust response to the uncertainty in demand, even with high fluctuations. In the third part, we propose a \textit{two-echelon vehicle routing problem} (2E-VRP) under consideration of a heterogeneous fleet of vehicles and different customer types. In our model, unlike the previous studies in the literature, not only do the large vehicles visit the pre-assigned points, called satellites, to refill the smaller vehicles, but they also deliver items to the customers. On the other hand, smaller vehicles are responsible for the customers with small size demands and can get refilled whether at the depots or satellites. We propose a branch-and-price algorithm as the solution approach and obtain promising results in comprehensive numerical studies that prove its versatility.Bu tezde, büyük ölçekli optimizasyon yöntemlerinin üç farklı uygulamasını ele alıyoruz. İlk problemde, belirsiz talep altında kan bağışı terzilik problemine odaklanıyoruz. İkincisinde, esnek imalat sistemleri ve tipik imalat makinelerinden oluşan hibrit imalat için bir model öneriyoruz. Sonuncusunda, bakkalların son mil teslimatı için iki kademeli bir araç rotalama problemini ele alıyoruz. Tezin ilk bölümünde, çok bileşenli aferezi benimseyen ve burada-şimdi ve bekle-gör donörleri olarak donör havuzu bölümlemesini kullanan kan bağışı yönetimi probleminin stokastik senaryo tabanlı yeniden formüle edilmesini öneriyoruz. Bağış havuzu segmentasyonu, geleneksel bağış yaklaşımından daha esnek bağış programları sağlar çünkü bekle ve gör bağışçılar bağış programlarını zaman içinde gerçekleşen talep değerlerine göre ayarlayabilirler. Gerçekçi olarak boyutlandırılmış örnekler için ilgili çok aşamalı stokastik bağış uyarlama problemini çözmek için bir sütun oluşturma yaklaşımı öneriyoruz. İkinci bölümde, düzenli talebi karşılamak için tipik olarak tahsis edilmiş makineler ve ani talebi karşılamak için esnek bir üretim sistemi ile esnek/hibrit üretim üretim ayarı ele alınmaktadır. Talepteki belirsizliği senaryo tabanlı bir yaklaşım kullanarak modelliyoruz ve işletmenin standart üretimin maliyet etkinliğinden ve esnek üretim sistemlerinin yanıt verebilirliğinden yararlanarak burada ve şimdi ve bekle-gör kararları almasına izin veriyoruz. Çözüm yaklaşımı olarak sütun oluşturma tabanlı bir algoritma öneriyoruz. Hesaplamalı analizimiz, bu hibrit üretim ayarının, yüksek dalgalanmalarda bile talepteki belirsizliğe son derece sağlam yanıt verdiğini gösteriyor.\\ Üçüncü bölümde, heterojen bir araç filosu ve farklı müşteri tipleri göz önünde bulundurularak \textit{iki kademeli bir araç rotalama problemi} (2K-ARP) önerilmiştir. Modelimizde literatürdeki önceki çalışmalardan farklı olarak büyük araçlar, daha küçük araçları doldurmak için uydu adı verilen önceden belirlenmiş noktaları ziyaret etmekle kalmaz, aynı zamanda müşterilere ürün teslim eder. Daha küçük araçlar ise, küçük boyutlu talepleri olan müşterilerin sorumluluğundadır ve ister depolarda ister uydularda dolum yapabilmektedir. Çözüm yaklaşımı olarak bir dal-ve-fiyat-kes algoritması öneriyoruz ve çok yönlülüğünü kanıtlayan kapsamlı sayısal çalışmalarda umut verici sonuçlar elde ediyoruz