CBL - Campus del Baix Llobregat

Projecte llegit

Títol: Image impainting using deep neural networks


Estudiants que han llegit aquest projecte:


Director: TARRÉS RUIZ, FRANCESC

Departament: TSC

Títol: Image impainting using deep neural networks

Data inici oferta: 26-01-2023     Data finalització oferta: 26-09-2023



Estudis d'assignació del projecte:
    MU MASTEAM 2015
Tipus: Individual
 
Lloc de realització: EETAC
 
Paraules clau:
image processing, image impainting, neural networks, machine learning, self-supervised neural networks, generative adversarial networks
 
Descripció del contingut i pla d'activitats:
The project will study and implement a deep learning strategy for implementing image



impainting, that is, create blocks of missing information in images. The system will be



based on a generative adversarial model and trained using a selfsupervised strategy. The



student will analyze different implementations and implement one of them using Pytorch



or Keras.
 
Overview (resum en anglès):
Nowadays, virtual markets are increasingly available and seek to connect shoppers with products. Due to the high turnover of products in a physical market, it is very likely to find relevant differences between products in the physical and digital markets.

This paper proposes the use of image inpainting using deep neural networks to solve this problem. It is proposed to use the approach performed by [1] based on generative adversarial networks as they are one of the most inventive and promising architectures. Through the experiments performed, it has been possible to prove that using this method it is possible to train models that produce realistic terminations of products that have been eliminated or that are to be replaced. We have also made a comparison with another interesting approach that had shown good results in the task of content generation in arbitrary zones.


© CBLTIC Campus del Baix Llobregat - UPC