Filters








4,345 Hits in 1.7 sec

Review with care

Adriana L. Romero-Olivares
2019 Science  
Keep in mind that writing in a second language is even harder. j Adriana L. Romero-Olivares is a postdoc at the University of New Hampshire in Durham.  ...  Adriana L. Romero-Olivares is a postdoc at the University of New Hampshire in Durham. Send your career story to SciCareerEditor@aaas.org.  ... 
doi:10.1126/science.366.6461.146 pmid:31604317 fatcat:f3a4c53gk5a7vl4dbvmotpiwom

Revisiting Hotels-50K and Hotel-ID [article]

Aarash Feizi, Arantxa Casanova, Adriana Romero-Soriano, Reihaneh Rabbany
2022 arXiv   pre-print
In this paper, we propose revisited versions for two recent hotel recognition datasets: Hotels50K and Hotel-ID. The revisited versions provide evaluation setups with different levels of difficulty to better align with the intended real-world application, i.e. countering human trafficking. Real-world scenarios involve hotels and locations that are not captured in the current data sets, therefore it is important to consider evaluation settings where classes are truly unseen. We test this setup
more » ... ng multiple state-of-the-art image retrieval models and show that as expected, the models' performances decrease as the evaluation gets closer to the real-world unseen settings. The rankings of the best performing models also change across the different evaluation settings, which further motivates using the proposed revisited datasets.
arXiv:2207.10200v1 fatcat:ac6cb5epxjbpddnb7palg5bpbu

Análisis de género en los factores relacionados con las caídas hospitalarias

Roberto Lara-Romero, María Victoria Ruiz-Romero, Aurelio Cayuela-Dominguez, Adriana Rivera-Sequeiros
2015 Zenodo  
Las caídas son un importante factor de salud pública, siendo el tipo de accidente más frecuente y letal entre las personas mayores. Algunos estudios han mostrado un incremento de las caídas a medida que aumenta la edad, mermando el nivel de dependencia de estas personas y complicando sus cuidados en salud. Como la mayoría de síndromes geriátricos, las caídas tienen una etiología mulitifactorial y la identificación de sus factores de riesgo en un 95% de los casos puede realizarse mediante la
more » ... oria clínica y la exploración física. El ámbito hospitalario tiene ciertas peculiaridades frente al entorno domiciliario, y nos planteamos conocer cómo se producen las caídas en el hospital y cuáles son los factores que intervienen teniendo en cuenta característica de género. Utilizamos un registro de caídas informático completado por personal de enfermería. Los datos recogidos durante 3 años fueron 92 caídas registradas de 89 pacientes, cuya edad media fue de 69 años (SD:13,8) y la relación hombre/mujer fue del 1,2. Tras los resultados identificamos una lista de factores del entorno que se relacionaron con las caídas, además de factores intrínsecos de la persona. Además se identificó la menor solicitud de ayuda en hombres como una diferencia significativa de género en factores predisponentes (p<0,05). La implantación de estas medidas de prevención identificadas, de forma sistemática y ajustadas al riesgo del paciente pretenden reducir este evento adverso hospitalario, siendo un punto de partida para próximas investigaciones.
doi:10.5281/zenodo.3534354 fatcat:n53mk2qn5batzdw3hetco5g2gi

Instance-Conditioned GAN [article]

Arantxa Casanova, Marlène Careil, Jakob Verbeek, Michal Drozdzal, Adriana Romero-Soriano
2021 arXiv   pre-print
Generative Adversarial Networks (GANs) can generate near photo realistic images in narrow domains such as human faces. Yet, modeling complex distributions of datasets such as ImageNet and COCO-Stuff remains challenging in unconditional settings. In this paper, we take inspiration from kernel density estimation techniques and introduce a non-parametric approach to modeling distributions of complex datasets. We partition the data manifold into a mixture of overlapping neighborhoods described by a
more » ... datapoint and its nearest neighbors, and introduce a model, called instance-conditioned GAN (IC-GAN), which learns the distribution around each datapoint. Experimental results on ImageNet and COCO-Stuff show that IC-GAN significantly improves over unconditional models and unsupervised data partitioning baselines. Moreover, we show that IC-GAN can effortlessly transfer to datasets not seen during training by simply changing the conditioning instances, and still generate realistic images. Finally, we extend IC-GAN to the class-conditional case and show semantically controllable generation and competitive quantitative results on ImageNet; while improving over BigGAN on ImageNet-LT. Code and trained models to reproduce the reported results are available at https://github.com/facebookresearch/ic_gan.
arXiv:2109.05070v2 fatcat:a7myozdljfhibgmkdcqn6bo6xa

Parameter Prediction for Unseen Deep Architectures [article]

Boris Knyazev, Michal Drozdzal, Graham W. Taylor, Adriana Romero-Soriano
2021 arXiv   pre-print
Deep learning has been successful in automating the design of features in machine learning pipelines. However, the algorithms optimizing neural network parameters remain largely hand-designed and computationally inefficient. We study if we can use deep learning to directly predict these parameters by exploiting the past knowledge of training other networks. We introduce a large-scale dataset of diverse computational graphs of neural architectures - DeepNets-1M - and use it to explore parameter
more » ... rediction on CIFAR-10 and ImageNet. By leveraging advances in graph neural networks, we propose a hypernetwork that can predict performant parameters in a single forward pass taking a fraction of a second, even on a CPU. The proposed model achieves surprisingly good performance on unseen and diverse networks. For example, it is able to predict all 24 million parameters of a ResNet-50 achieving a 60% accuracy on CIFAR-10. On ImageNet, top-5 accuracy of some of our networks approaches 50%. Our task along with the model and results can potentially lead to a new, more computationally efficient paradigm of training networks. Our model also learns a strong representation of neural architectures enabling their analysis.
arXiv:2110.13100v1 fatcat:6rahyzxow5dtvp4vsypi6czbj4

Introduction and Acknowledgments

Sarah S.V. Cantor, Federica Di Blasio, Adriana Guarro Romero
2019 Carte italiane  
Cantor, Editor-in-Chief Adriana Guarro Romero, Editor-in-Chief Federica Di Blasio, Managing Editor  ... 
doi:10.5070/c9121046097 fatcat:njbmg6pwnbbmnnpmsvzwh5dcia

Reseñas de libros

Adriana Suárez Sánchez, Alejandra Martínez Romero, Hugo Alberto Figueroa Alcántara
2014 Biblioteca Universitaria  
Adriana Suárez Sánchez Programa del Doctorado en Bibliotecología y Estudios de la Información L a Fundación Calouste Gulbenkian (institución privada portuguesa que se dedica a promover el arte, la caridad  ...  social una buena guía para comprender el desarrollo de estas disciplinas, y es un referente para comprender la metodología de la investigación en ciencias sociales en la actualidad. a Alejandra Martínez Romero  ... 
doi:10.22201/dgb.0187750xp.2013.1.26 fatcat:myvjtvf5bbfnlc3ga6sysufn7m

GEOMetrics: Exploiting Geometric Structure for Graph-Encoded Objects [article]

Edward J. Smith, Scott Fujimoto, Adriana Romero, David Meger
2019 arXiv   pre-print
Mesh models are a promising approach for encoding the structure of 3D objects. Current mesh reconstruction systems predict uniformly distributed vertex locations of a predetermined graph through a series of graph convolutions, leading to compromises with respect to performance or resolution. In this paper, we argue that the graph representation of geometric objects allows for additional structure, which should be leveraged for enhanced reconstruction. Thus, we propose a system which properly
more » ... efits from the advantages of the geometric structure of graph encoded objects by introducing (1) a graph convolutional update preserving vertex information; (2) an adaptive splitting heuristic allowing detail to emerge; and (3) a training objective operating both on the local surfaces defined by vertices as well as the global structure defined by the mesh. Our proposed method is evaluated on the task of 3D object reconstruction from images with the ShapeNet dataset, where we demonstrate state of the art performance, both visually and numerically, while having far smaller space requirements by generating adaptive meshes
arXiv:1901.11461v1 fatcat:qtmpheqtxjbjnlwy6foashqh7q

Nuevos territorios urbanos: consideraciones de la espacialidad contemporánea

Adriana I. Olivares González, Daniel González Romero
2004 Urbano  
Al observarse las primeras implicaciones espaciales del modelo neoliberal, diversos investigadores predijeron la crisis de las grandes concentraciones urbanas e incluso el fin de las grandes ciudades. En efecto el importante desarrollo de los medios e infraestructuras de transporte y sobre todo el avance de las tecnologías de la Información y la Comunicación (TIC), han constituido desde entonces el soporte que impulso a una dinámica espacial con tendencia a la descentralización, tanto de las actividades económicas como de la población.
doaj:5ab5d6ad4b454f82b6228ebbbec20a89 fatcat:f25fzzxhjfagzn3kjhvjemhs4m

No more meta-parameter tuning in unsupervised sparse feature learning [article]

Adriana Romero, Petia Radeva, Carlo Gatta
2014 arXiv   pre-print
We propose a meta-parameter free, off-the-shelf, simple and fast unsupervised feature learning algorithm, which exploits a new way of optimizing for sparsity. Experiments on STL-10 show that the method presents state-of-the-art performance and provides discriminative features that generalize well.
arXiv:1402.5766v1 fatcat:cmncgzcwzfgijl6x5xz2c2kkbq

Clubes de lectura: estrategia pedagógica para la generación de una cultura lectora

Maria Edilse Álvarez Romero, Claudia Patricia Castañeda Ariza, Leiby Adriana García Romero
2019 Educación y ciencia  
La escuela debe promover la creación de estos ambientes de aprendizaje, y María Edilse Álvarez Romero -Claudia Patricia Castañeda Ariza -Leiby Adriana García Romero Educación y Ciencia -Núm 22.  ...  Año 2019 • Pág. 375-386 María Edilse Álvarez Romero -Claudia Patricia Castañeda Ariza -Leiby Adriana García Romero Maratones de lectura: mediante la maratón de lectura, se logró desarrollar los  ... 
doi:10.19053/0120-7105.eyc.2019.22.e10058 fatcat:ws7rdusywfcj3avhqpkmxcmpfm

ANNULOHYPOXYLON (HYPOXYLACEAE) SPECIES FROM ARGENTINA

Esteban B. Sir, Erik Kuhnert, Adriana I. Hladki, Andrea I. Romero
2018 Darwiniana  
Romero, A. nitens (Ces.) Y.M. Ju, J.D. Rogers & H.M. Hsieh, A. subeffusum (Speg.) Hladki & A.I. Romero, A. stygium (Lév.) Y.M. Ju, J.D. Rogers & H. M. Hsieh., A. substygium (Y. M. Ju, J.D.  ...  In South America, 19 species are recognized (Cruz & Cortez, 2016; Hladki & Romero, 2009a, b; Ju & Rogers, 1996; Kuhnert et al., 2017) (Hladki & Romero 2009a, b; Kuhnert et al., 2017; Daranagama et  ... 
doi:10.14522/darwiniana.2018.61.777 fatcat:vpkopuwqwve7xc52tmyjdsrbc4

Graph Attention Networks [article]

Petar Veličković, Guillem Cucurull, Arantxa Casanova, Adriana Romero, Pietro Liò, Yoshua Bengio
2018 arXiv   pre-print
We present graph attention networks (GATs), novel neural network architectures that operate on graph-structured data, leveraging masked self-attentional layers to address the shortcomings of prior methods based on graph convolutions or their approximations. By stacking layers in which nodes are able to attend over their neighborhoods' features, we enable (implicitly) specifying different weights to different nodes in a neighborhood, without requiring any kind of costly matrix operation (such as
more » ... inversion) or depending on knowing the graph structure upfront. In this way, we address several key challenges of spectral-based graph neural networks simultaneously, and make our model readily applicable to inductive as well as transductive problems. Our GAT models have achieved or matched state-of-the-art results across four established transductive and inductive graph benchmarks: the Cora, Citeseer and Pubmed citation network datasets, as well as a protein-protein interaction dataset (wherein test graphs remain unseen during training).
arXiv:1710.10903v3 fatcat:6xbvqpuxsjfuhfo7vqcdjmefgq

Learning to adapt class-specific features across domains for semantic segmentation [article]

Mikel Menta, Adriana Romero, Joost van de Weijer
2020 arXiv   pre-print
And finally, I would like to thank my supervisor Adriana, first of all for offering me the opportunity of joining the AI community of Montreal and letting me work with her in this project.  ... 
arXiv:2001.08311v1 fatcat:xeqowbv2bzb5vaitwfuqszlj4a

Julián David Romero Torres. "A la lucha he venido". La campaña electoral de 1930 en Colombia. Bogotá: Universidad del Rosario, 2018

Adriana Rodríguez
2020 Historia Caribe  
Adriana Rodríguez Franco Universidad del Tolima arodriguezfr@ut.edu.co arodriguezfr@gmail.com Julián David Romero Torres. "A la lucha he venido".  ...  Restrepo), se sobreentiende que la obra de Romero abordaría la cuestión del republicanismo.  ... 
doi:10.15648/hc.36.2020.13 fatcat:d6xkoqkinbeozbqun7pn6opprq
« Previous Showing results 1 — 15 out of 4,345 results