Facultad de Matemáticas e Informática

 La Facultad 
Conoce la Facultad
Bienvenida del decano
Indicadores y ránquings
Profesorado
Usos lingüísticos
Instalaciones de la Facultad
Lecciones inaugurales
Vivir en Barcelona

Organización y estructura
Órganos de gobierno
Decanato
Consejos de Estudios
Coordinación de los másteres
Administración
	Administración del Centro
Oficina de Investigación
Oficina de Asuntos Generales
Servicios TIC

Acuerdos

Actividades Facultad
Premios y ayudas
Departamento
Secretaría
Comisión de Igualdad
Sistema de calidad
Presentación
Sistema de aseguramiento interno de la calidad
	Presentación del SAIQU – Matemáticas e Informática
Política y objetivos de calidad
Plan director
Comisión de Calidad
Manual de calidad
Gestión de procesos del SAIQU
Revisión y mejora
Certificación del SAIQU de la Facultad

Verificación, Seguimento, Modificación, Acreditación
	Presentación
Memorias de verificación
Informe de seguimiento de Centro
Acreditación

Datos e indicadores
	Cuadro de mando
Indicadores académicos y de satisfacción: Espacio VSMA y Business intelligence
Satisfacción de las partes interesadas
Inserción laboral de titulados
Otros indicadores

                Agenda
            
                            Categoria:
                            
                                    CVUB Seminar: Overcoming Scalability and Efficiency Challenges in Fine-Grained Classification with Few-Shot Learning
 Seminario | 03/11/2023  
 A càrrec de Petia Radeva  
 Organización: Artificial Intelligence and Biomedical Applications  

CVUB Seminar: Overcoming Scalability and Efficiency Challenges in Fine-Grained Classification with Few-Shot Learning   

                                        Seminario
                                    
                                            Fecha:                                        

                                            03/11/2023
                                        
                                            Horario:                                        

                                            15:00
                                        
                                            Lugar:                                        

                                            Hybrid (Teams / IMUB)
                                        
                                                A cargo de:                                            

                                                Petia Radeva
                                            
                                            Organización:                                        

                                            Artificial Intelligence and Biomedical Applications
                                        
                                                Observaciones:                                            

                                                Fine-grained classification tasks, which require discerning between highly similar categories, pose significant challenges in scalability and efficiency. Traditionally, these tasks demand extensive labelled datasets for training complex models, a process that becomes impractical when considering the cost, time, and expertise needed to collect and label data. Additionally, the computational burden of training and updating these large-scale models is non-trivial, especially in dynamic environments where categories evolve rapidly. Few-shot learning emerges as a solution to these issues. It reduces the reliance on large datasets by enabling models to learn and make accurate predictions from a limited number of examples.
                                            
                                Salir
                            
                                    Sessió informativa sobre la convocatòria de mobilitat internacional (Erasmus) 2024-2025
 Otros | 20/10/2023  
 A càrrec de Oficina de Relacions Internacionals  
 Organización: Oficina de Relacions Internacionals  

Sessió informativa sobre la convocatòria de mobilitat internacional (Erasmus) 2024-2025   

                                        Otros
                                    
                                            Fecha:                                        

                                            20/10/2023
                                        
                                            Horario:                                        

                                            13h.
                                        
                                            Lugar:                                        

                                            Aula B3
                                        
                                                A cargo de:                                            

                                                Oficina de Relacions Internacionals
                                            
                                            Organización:                                        

                                            Oficina de Relacions Internacionals
                                        
                                                Observaciones:                                            

                                                Per a més informació, podeu consultar l'enllaç següent: http://www.ub.edu/uri/estudiantsUB/erasmus.htm
                                            
                                Salir
                            
                                    CVUB Seminar: Dining on Details: LLM-Guided Expert Networks for Fine-Grained Food Recognition
 Seminario | 20/10/2023  
 A càrrec de Petia Radeva  
 Organización: Computer Vision and Machine Learning Group  

CVUB Seminar: Dining on Details: LLM-Guided Expert Networks for Fine-Grained Food Recognition   

                                        Seminario
                                    
                                            Fecha:                                        

                                            20/10/2023
                                        
                                            Horario:                                        

                                            15 h
                                        
                                            Lugar:                                        

                                            Hybrid (Teams / IMUB)
                                        
                                                A cargo de:                                            

                                                Petia Radeva
                                            
                                            Organización:                                        

                                            Computer Vision and Machine Learning Group
                                        
                                                Observaciones:                                            

                                                Dining on Details: LLM-Guided Expert Networks for Fine-Grained Food Recognition\" introduces a novel approach in the field of fine-grained food recognition. The method leverages large language models and multi-modal information to identify and classify groups of highly similar dishes. The innovative methodology, developed through an end-to-end multi-task learning process, enhances accuracy in food classification. Notably, the method\'s universal compatibility allows it to be seamlessly integrated into any existing classification architecture, paving the way for versatile future research endeavours. The comprehensive validation of this method across various food datasets and backbones reveals significant performance gains, and it achieves state-of-the-art results on the Food-101 dataset.
                                            
                                Salir
                            
                                    CVUB Seminar: Developing Edge AI Computer Vision for Smart Poultry Farms Using Deep Learning and HPC
 Seminario | 13/10/2023  
 A càrrec de Petia Radeva  
 Organización: Computer Vision and Machine Learning Group  

CVUB Seminar: Developing Edge AI Computer Vision for Smart Poultry Farms Using Deep Learning and HPC   

                                        Seminario
                                    
                                            Fecha:                                        

                                            13/10/2023
                                        
                                            Horario:                                        

                                            15 h 
                                        
                                            Lugar:                                        

                                            Hybrid (Teams / IMUB)
                                        
                                                A cargo de:                                            

                                                Petia Radeva
                                            
                                            Organización:                                        

                                            Computer Vision and Machine Learning Group
                                        
                                                Observaciones:                                            

                                                This research describes high-performance computing (HPC) and deep learning to create prediction models that could be deployed on edge AI devices equipped with cameras and installed in poultry farms. The main idea is to leverage an existing IoT farming platform and use HPC offline to run deep learning to train the models for object detection and object segmentation, where the objects are chickens in images taken on a farm. The models can be ported from HPC to edge AI devices to create a new type of computer vision kit to enhance the existing digital poultry farm platform. Such new sensors enable implementing functions such as counting chickens, detecting dead chickens, and even assessing their weight or detecting uneven growth. These functions combined with the monitoring of environmental parameters, could enable early disease detection and improve the decision-making process.
                                            
                                Salir
                            
                                    Lògica: eina i fonament
 Otros | 4, 11, 18 i 25 d'octubre  
 A càrrec de Estudiants del Màster en Lògica Pura i Aplicada  
 Organización: Màster en Lògica Pura i Aplicada  

Lògica: eina i fonament   

                                        Otros
                                    
                                            Fecha:                                        

                                            4, 11, 18 i 25 d'octubre
                                        
                                            Horario:                                        

                                            De 12 a 14 h
                                        
                                            Lugar:                                        

                                            Aula B6
                                        
                                                A cargo de:                                            

                                                Estudiants del Màster en Lògica Pura i Aplicada
                                            
                                            Organización:                                        

                                            Màster en Lògica Pura i Aplicada
                                        
                                Salir
                            
                                    Acte d’inauguració del curs 2023-24
 Otros | 27/09/2023  
 A càrrec de Maite López Sánchez  
 Organización: Facultat de Matemàtiques i Informàtica de la Universitat de Barcelona  

Acte d’inauguració del curs 2023-24   

                                        Otros
                                    
                                            Fecha:                                        

                                            27/09/2023
                                        
                                            Horario:                                        

                                            12.15
                                        
                                            Lugar:                                        

                                            Aula Magna
                                        
                                                A cargo de:                                            

                                                Maite López Sánchez
                                            
                                            Organización:                                        

                                            Facultat de Matemàtiques i Informàtica de la Universitat de Barcelona
                                        
                                                Observaciones:                                            

                                                En el transcurs de l'acte es lliurarà el Distinguished Paper Award Collectanea Mathematica 2022. 

Més informació
                                            
                                                Enlace web
                                            
                                Salir
                            
                                    CVUB Seminar: MomentsNeRF: Incorporating Orthogonal Moments in Convolutional Neural Networks for One or Few-Shot Neural Rendering.
 Seminario | 22/09/2023  
 A càrrec de Petia Radeva  
 Organización: Computer Vision and Machine Learning Group  

CVUB Seminar: MomentsNeRF: Incorporating Orthogonal Moments in Convolutional Neural Networks for One or Few-Shot Neural Rendering.   

                                        Seminario
                                    
                                            Fecha:                                        

                                            22/09/2023
                                        
                                            Horario:                                        

                                            15:00
                                        
                                            Lugar:                                        

                                            Hybrid (Teams / IMUB)
                                        
                                                A cargo de:                                            

                                                Petia Radeva
                                            
                                            Organización:                                        

                                            Computer Vision and Machine Learning Group
                                        
                                                Observaciones:                                            

                                                We propose MomentsNeRF, a one and a few-shot neural rendering framework that predicts a neural representation of a scene using Orthogonal Moments. Our architecture offers a transfer learning method to pre-train on multi-scenes and incorporate a per-scene optimization at test time using one or a few images. Our method leverages transfer learning, which learns from features extracted from orthogonal Gabor and Zernike moments. Our approach shows a better performance in synthesizing the scene details in terms of complex texture and shape capturing, noise reduction, artifact elimination, and completing the missing parts compared to the recent one and few-shot neural rendering frameworks. We conduct extensive experiments on real scenes from the DTU dataset. MomentsNeRF improves the existing approaches by 3.39 dB PSNR, 11.1% SSIM, 17.9% LPIPS, and 8.3% DISTS metrics. Moreover, MomentsNeRF achieves SoTA performance for both novel view synthesis and single-image 3D view reconstruction.
                                            
                                Salir
                            
                                    Simposi d’Administració Pública i Proposta de la UE per a un Reglament d’Intel·ligència Artificial
 Otros | 18 i 19 de setembre del 2023 (inscripcions fins al 14 de juliol i acceptacions el dia 31 de juliol)  
 A càrrec de Univeristat de Barcelona, Uniersitat Oberta de Catalunya, Ministeri de Ciència i Innovació, Agència Estatal d'Investigació  
 Organización: Univeristat de Barcelona, Uniersitat Oberta de Catalunya, Ministeri de Ciència i Innovació, Agència Estatal d'Investigació  

Simposi d’Administració Pública i Proposta de la UE per a un Reglament d’Intel·ligència Artificial   

                                        Otros
                                    
                                            Fecha:                                        

                                            18 i 19 de setembre del 2023 (inscripcions fins al 14 de juliol i acceptacions el dia 31 de juliol)
                                        
                                            Horario:                                        

                                            Lugar:                                        

                                            Facultat de Dret (Universitat de Barcelona)
                                        
                                                A cargo de:                                            

                                                Univeristat de Barcelona, Uniersitat Oberta de Catalunya, Ministeri de Ciència i Innovació, Agència Estatal d'Investigació
                                            
                                            Organización:                                        

                                            Univeristat de Barcelona, Uniersitat Oberta de Catalunya, Ministeri de Ciència i Innovació, Agència Estatal d'Investigació
                                        
                                                Enlace web
                                            
                                Salir
                            
                                    Seminari «Getting deeper into learning with noisy labeling»
 Seminario | 21/07/2023  
 A càrrec de Petia Radeva  
 Organización: Computer Vision and Machine Learning Group  

Seminari «Getting deeper into learning with noisy labeling»   

                                        Seminario
                                    
                                            Fecha:                                        

                                            21/07/2023
                                        
                                            Horario:                                        

                                            15:00
                                        
                                            Lugar:                                        

                                            Hybrid (Teams / IMUB)
                                        
                                                A cargo de:                                            

                                                Petia Radeva
                                            
                                            Organización:                                        

                                            Computer Vision and Machine Learning Group
                                        
                                                Observaciones:                                            

                                                In the field of deep learning, a large number of correctly labelled samples are essential for model training. However, in practical applications, labelling data requires high labelling costs. At the same time, label noise cannot be avoided during the labelling process. In this seminar, we will discuss the origins and impacts of learning with noisy label problems, explore existing learning with noisy label methods and discuss the development of future learning with noisy label methods.
                                            
                                Salir
                            
                                    Seminari «CVUB Seminar: Phenotype prediction using ML and graph neural network in Alzheimer’s disease»
 Seminario | 14/07/2023  
 A càrrec de Petia Radeva  
 Organización: Computer Vision and Machine Learning Group  

Seminari «CVUB Seminar: Phenotype prediction using ML and graph neural network in Alzheimer’s disease»   

                                        Seminario
                                    
                                            Fecha:                                        

                                            14/07/2023
                                        
                                            Horario:                                        

                                            15:00h
                                        
                                            Lugar:                                        

                                            Hybrid (Teams / IMUB)
                                        
                                                A cargo de:                                            

                                                Petia Radeva
                                            
                                            Organización:                                        

                                            Computer Vision and Machine Learning Group
                                        
                                                Observaciones:                                            

                                                In this comprehensive research study, we explored the association between genetic markers and the prediction of Alzheimer\'s disease (AD). Our main objective was to gain insights into the complex interplay between genetic variations and the likelihood of developing this neurodegenerative disorder. By employing advanced Machine and Deep learning (ML and DL) techniques, we aimed to unravel the underlying mechanisms and identify key factors influencing the prediction of AD. The study focused on analysing genetic data and utilizing ML models for phenotype prediction. Traditional methods based on logistic regression models were compared to various ML algorithms, such as Random Forest, XGBoost, LightGBM and also novel DL approaches using Graph Neural Networks (GNNs).
                                            
                                Salir
                            
chevron_left
1
…
3
4
5
6
7
…
18
chevron_right

                dialpad
                                Portales
                            
                            Món UB
                            Campus Virtual
                            Alumni UB
                            Intranet (PDI i PAS)
                        
                        Formulario de actividades