Experience

Machine Learning Reply

• Italy
• IT Services and IT Consulting
• 1 - 100 Employee

• Cloud, DevOps and Machine Learning Engineer

  • Sep 2022 - Present



AImageLab

• Italy
• Research Services
• 1 - 100 Employee

• Studente tirocinante

  • Sep 2021 - Apr 2022

  The work done during my thesis led to the creation of an architecture called DBCT, a Transformer-based architecture for virtual try-on. This innovative system can jointly exploit cross-modal information of in-shop garments and reference models at generation time. Through experimental analyses on the VITON dataset and on Dress Code, a dataset collected by AImageLab with the collaboration of YOOX NET-A-PORTER Group, we have demonstrated the effectiveness of the proposed solution in comparison to both standard pure convolutional approaches and previous Transformer-based proposals for the task. I am proud to announce that this project has been accepted in the CVPR workshop. Show less