336 research outputs found
Cyclic Behaviour of Thin-Walled Pre-cracked HPFRC in Bending
No full textApplications of High-Performance Fibre-Reinforced Concrete (HPFRC) are gaining popularity both thanks to its enhanced mechanical performance and its capability of guaranteeing pleasant finishing that makes this type of materials a good solution for thin elements like façade panels. These kinds of elements are often subjected to cyclic loads throughout their service life. The impact of cyclic loads on material properties is significant and can potentially lead to fatigue failures, particularly in situations in which the elements have experienced cracks during their service life. The research presented here is oriented to studying the cyclic bending behaviour of thin-walled elements made of one specific HPFRC material, also considering different level of pre-cracking that refers both to serviceability limit state and to ultimate limit state condition. A set of four-point bending tests were performed on 35 mm thick samples applying a pulsating load with a cycle frequency of 1.3 Hz and considering different load range typical of the serviceability limit state. The pre-crack levels considered correspond to 0.5, 1.5 and 2.5 mm of global crack opening measured astride the constant bending moment region. The crack evolution has been measured during the proceeding of the cycles and, in the cases in which the samples did not experience failure before 100.000 cycles, a monotonic test was performed in order to compare the structural performance after cycles with that in pristine condition
Harnessing Foundation Models for Image Anonymization
No full textTraditional deep learning pipelines involve multiple intricate steps, from data acquisition to model training, finetuning, and deployment. However, recent advancements in foundation models, particularly in text-to-image generation, offer a paradigm shift in addressing tasks without the need for these conventional processes. In this paper, we explore how foundation models can be leveraged to solve tasks, specifically focusing on anonymization, without the requirement for training or fine-tuning. By bypassing traditional pipelines, we demonstrate the efficiency and effectiveness of this approach in achieving anonymization objectives directly from the foundation model’s inherent knowledge. Our findings underscore the transformative potential of foundation models in simplifying and accelerating deep learning tasks, paving the way for novel applications in various domains
Explaining Concept Drift via Neuro-Symbolic Rules
No full textConcept drift in machine learning refers to changes in the underlying data distribution over time, which can lead to a degradation in the performance of predictive models. Although many methods have been proposed to detect and adapt to concept drift, effective methods to explain it in a human-understandable manner remain lacking. To address this, we propose the use of neuro-symbolic rules to explain the reason for drift.
We applied recent rule extraction methods to convolutional neural networks (CNNs) to shed light on the model's internal behavior and promote interpretability of the outputs, while also proposing two novel automated approaches for semantic kernel labeling.
We conducted preliminary experiments to assess the applicability and effectiveness of these rules in explaining concept drift, and the efficacy of the kernel labeling strategies.
Under the optimality assumption, our method was able to extract rules that can facilitate the identification of the causes of drift, through either rule inspection or antecedents activation frequencies analysis.
Moreover, the proposed strategies for kernel labeling offer a more reliable and scalable alternatives to the state-of-the-art solutions
Size effect in bending of High Performance Fiber Reinforced Concrete (HPFRC) thin slabs
No full textThe use of HPFRC, due to its characteristics of ductility, flexural strength, and pleasant finishing, has been extended to the construction of thin elements such as façades. However, the reliability of mechanical performance of this composite, determined on small samples for designing large elements, can be debated; especially if no traditional reinforcement is introduced and both resistance and ductility of the member response rely solely on the contribution of fibers. The experimental campaign, proposed hereby, refers to a HPFRC with a volume fraction of 2.5% of hooked-end steel fibers and analyses 3.5 cm thick elements. Several four-point bending tests have been performed on samples varying in length (60 to 180 cm) and width (15 to 45 cm). The paper presents the main results of this investigation aimed at providing useful information on the size effect in bending of a HPFRC
For a semiotic AI: Bridging computer vision and visual semiotics for computational observation of large scale facial image archives
No full textSocial networks are creating a digital world in which the cognitive, emotional, and pragmatic value of the imagery of human faces and bodies is arguably changing. However, researchers in the digital humanities are often ill-equipped to study these phenomena at scale. This work presents FRESCO (Face Representation in E-Societies through Computational Observation), a framework designed to explore the socio-cultural implications of images on social media platforms at scale. FRESCO deconstructs images into numerical and categorical variables using state-of-the-art computer vision techniques, aligning with the principles of visual semiotics. The framework analyzes images across three levels: the plastic level, encompassing fundamental visual features like lines and colors; the figurative level, representing specific entities or concepts; and the enunciation level, which focuses particularly on constructing the point of view of the spectator and observer. These levels are analyzed to discern deeper narrative layers within the imagery. Experimental validation confirms the reliability and utility of FRESCO, and we assess its consistency and precision across two public datasets. Subsequently, we introduce the FRESCO score, a metric derived from the framework's output that serves as a reliable measure of similarity in image content
A Method for Detecting Structural Breaks and an Application to the Turkish Stock Market
No full textSidika Basci; Erdem Basc
Factors Related to Pulse Wave Velocity and Augmentation Index in Chronic Hemodialysis Patients
No full textBackground: Augmentation index (AIx) and pulse wave velocity (PWV) are early markers of atherosclerotic vascular changes and also have been shown to be predictive of cardiovascular disease and total mortality. The aim of our study was to evaluate the relationship between PWV and AIx-HR75, which is the corrected form of AIx according to a heart rate of 75 beats/min, echocardiographic parameters and biochemical parameters in chronic hemodialysis (HD) patients. Subjects and methods: AIx-HR75 and PWV were measured in 556 HD patients by applanation tonometry using the SphygmoCor device. Results: The mean PWV and AIx-HR75 values of the study group were 10.2 +/- 2.4 and 28.4 +/- 10.2 m/s. A positive correlation was found between PWV and AIx-HR75 (r = 0.214, p = 0.000). AIx-HR75 correlated with age (r = 0.093, p = 0.028), body surface area (BSA) (r = -0.194, p = 0.000), mean arterial pressure (MAP) (r = 0.335, p = 0.000), pulse pressure (PP) (r = 0.212, p = 0.000), cardiothoracic index (r = 0.155, p = 0.016), and presence of left ventricular hypertrophy (r = 0.152, p = 0.001). PWV correlated with MAP (r = 0.208, p = 0.000), PP (r = 0.098, r = 0.021), left ventricular mass (r = 0.105, p = 0.023), and predialysis sodium level (r = -0.105, p = 0.023). In the multivariate analyses, PWV was associated with MAP (t = 3.78, p = 0.000), presence of diabetes (t = 3.20, p = 0.001), and predialysis sodium level (t = -2.06, p = 0.040), and AIx-HR75 was associated with age (t = 2.48, p = 0.014), female sex (t = 3.98, p = 0.000), BSA (t = -2.15, p = 0.033), and MAP (t = 7.02, p = 0.000). Conclusion: There is a strong association between MAP and arterial stiffness parameters in HD patients. We feel that efficient control of blood pressure could lead to reduced arterial stiffness in HD patients
Biological reconstruction of the bone defects with free fibula flap after resections of extremity located bone tumors: Clinical and radiological short-term results
No full textBackground: Recently, limb salvage surgery is a preferred method in orthopedic oncology and extremity-located bone tumors treated by limb salvage surgery have a 90%-95% success rate. The aim of the study is functional and radiological evaluations of the undergone biological reconstruction with free fibula flap (FFF) after tumor surgery and the effects of the defect size on the functional results. Subjects and Methods: Between 2005 and 2010, 13 patients (7M/6F) who underwent limb salvage surgery for benign/malignant bone tumors were included in study. Diagnoses included five osteosarcomas, six Ewing's sarcomas, one high-grade chondrosarcoma, and one aneurysmal bone cyst. Diaphyseal and metaphyseal regions of femur (7), humerus (3), tibia (2), and radius (1) were reconstructed. FFF was combined with a strut femoral allograft in seven cases. Postoperatively, partial weight-bearing allowed at postoperative 3 month and increased gradually. The mean follow-up was 25 months (12-60) and evaluated by extremity function scoring of Musculoskeletal Tumor Society (MSTS) and radiologically. Results: On the 6(th) month, in 92.3% of patients (12/13), evident union, and on the 12(th) month, in all of the patients, evident union and bone flap hypertrophy were observed. The mean MSTS score was measured 77.58% (46.66-100). As the resection size increased, the MSTS scores were significantly decreased (P = 0.027); as the bone flap size increased, there were relatively low MSTS scores (P = 0.440). On the patients without bone flap hypertrophy on 6(th) month, the bone flap size was measured relatively higher (P = 0.069) and the operation duration was relatively higher (P = 0.100). As the operation duration increased, there were relatively lower MSTS scores (P = 0.062). In cases where allograft and VFG combined (7/9 patients) had higher MSTS scores than the ones, only FFF was used (P = 0.621). Conclusions: Limb salvage surgery improves the life quality without worsening the prognosis and is a method that should be preferred. The biologic reconstruction of the defects with FFF, following extremity located musculoskeletal tumor resections have positive effects on functional outcomes
- …
