University of Twente Student Theses

Login

Ensuring identity preservation for motion translation in faces using VICE-GAN

Oostveen, Ronan (2022) Ensuring identity preservation for motion translation in faces using VICE-GAN.

[img] PDF
1MB
Abstract:In this paper, we propose further improving the identity- preserving features of the VICE-GAN network. The VICE- GAN network is a network that generates a video of a face expressing different emotions than the input video while preserving the same face. We suggest that using a more robust encoder could achieve these improvements. Another encoder could improve upon identity preservation because the encoder proposed in the original paper performs poorly on faces that did not appear in the training set. Therefore using an encoder that performs better on facial feature ex- traction on unseen faces, such as FaceNet could also improve the accuracy of the VICE-GAN on unseen face models
Item Type:Essay (Bachelor)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Programme:Computer Science BSc (56964)
Link to this item:https://purl.utwente.nl/essays/93858
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page