Temporally-coherent video cartoonization

The automatic transformation of short background videos from real scenarios into others with a visually pleasing style, like those used in cartoons, holds application in various domains. These include animated films, video games, advertisements, and many other areas that involve visual content creat...

Full description

Bibliographic Details
Author: Rayo Hernandez, Gustavo Enrique
Format: master thesis
Publication Date:2024
Country:España
Institution:Universitat Politècnica de Catalunya (UPC)
Repository:UPCommons. Portal del coneixement obert de la UPC
Language:English
OAI Identifier:oai:upcommons.upc.edu:2117/410264
Online Access:https://hdl.handle.net/2117/410264
Access Level:Open access
Keyword:Video recording
Artificial intelligence
Video cartoonization
video-to-video translation
diffusion model
Stable Diffusion
ControlNet
EbSynth
Vídeo
Intel·ligència artificial
Àrees temàtiques de la UPC::Informàtica::Intel·ligència artificial
Description
Summary:The automatic transformation of short background videos from real scenarios into others with a visually pleasing style, like those used in cartoons, holds application in various domains. These include animated films, video games, advertisements, and many other areas that involve visual content creation. A method or tool that can perform this task, would inspire, facilitate, and streamline the work of artists and people who produce this type of content. This thesis proposes a method that integrates multiple components to translate short background videos into others that contain a particular style. We employ Stable Diffusion, a text-to-image diffusion model, along with other technologies like ControlNet to translate keyframes from the source video, ensuring content preservation. The style of the transformed keyframes is propagated to the rest of the frames using EbSynth to make the process faster and maintain the temporal coherence. We quantitatively assess content preservation and temporal coherence using CLIP-based metrics over a new dataset of videos translated into three distinct styles. The implementation of our method is publicly available at https://github.com/gustavorayo/video-to-cartoon