This document describes the work carried out during the internship period, lasting about three hundred hours, by the graduating student Trinca Nicolò at the company Fashion Box S.p.a. The primary goal of this internship is to conduct a feasibility study and implement a cutting-edge Generative Adversarial Network (GAN) capable of generating realistic images of clothing, starting from a collection of images. This process involves selecting appropriate GAN architectures and the right technologies to achieve the desired outcome. Additionally, a proof of concept (POC) was developed to synthesize images and explore the possibility of modifying and driving the generation through a text description.

This document describes the work carried out during the internship period, lasting about three hundred hours, by the graduating student Trinca Nicolò at the company Fashion Box S.p.a. The primary goal of this internship is to conduct a feasibility study and implement a cutting-edge Generative Adversarial Network (GAN) capable of generating realistic images of clothing, starting from a collection of images. This process involves selecting appropriate GAN architectures and the right technologies to achieve the desired outcome. Additionally, a proof of concept (POC) was developed to synthesize images and explore the possibility of modifying and driving the generation through a text description.

Deep Learning Techniques applied to Clothing Image Synthesis and Fashion Styling

TRINCA, NICOLO'
2022/2023

Abstract

This document describes the work carried out during the internship period, lasting about three hundred hours, by the graduating student Trinca Nicolò at the company Fashion Box S.p.a. The primary goal of this internship is to conduct a feasibility study and implement a cutting-edge Generative Adversarial Network (GAN) capable of generating realistic images of clothing, starting from a collection of images. This process involves selecting appropriate GAN architectures and the right technologies to achieve the desired outcome. Additionally, a proof of concept (POC) was developed to synthesize images and explore the possibility of modifying and driving the generation through a text description.
2022
Deep Learning Techniques applied to Clothing Image Synthesis and Fashion Styling
This document describes the work carried out during the internship period, lasting about three hundred hours, by the graduating student Trinca Nicolò at the company Fashion Box S.p.a. The primary goal of this internship is to conduct a feasibility study and implement a cutting-edge Generative Adversarial Network (GAN) capable of generating realistic images of clothing, starting from a collection of images. This process involves selecting appropriate GAN architectures and the right technologies to achieve the desired outcome. Additionally, a proof of concept (POC) was developed to synthesize images and explore the possibility of modifying and driving the generation through a text description.
StyleGan
CLIP
Fashion
GAN
Design
File in questo prodotto:
File Dimensione Formato  
Trinca_Nicolò.pdf

accesso riservato

Dimensione 28.53 MB
Formato Adobe PDF
28.53 MB Adobe PDF

The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.12608/52311