Real Face Video Animation Platform
This work addresses a problem for users needing high-quality anime-style facial animations, but it is incremental as it builds on existing models and datasets.
The authors tackled the lack of expressive power in facial video generation for exaggerated anime-style faces by proposing a real-time animation platform that converts real human faces to cartoon-style faces, achieving this through a user-friendly system built on Gradio with support for multiple models.
In recent years, facial video generation models have gained popularity. However, these models often lack expressive power when dealing with exaggerated anime-style faces due to the absence of high-quality anime-style face training sets. We propose a facial animation platform that enables real-time conversion from real human faces to cartoon-style faces, supporting multiple models. Built on the Gradio framework, our platform ensures excellent interactivity and user-friendliness. Users can input a real face video or image and select their desired cartoon style. The system will then automatically analyze facial features, execute necessary preprocessing, and invoke appropriate models to generate expressive anime-style faces. We employ a variety of models within our system to process the HDTF dataset, thereby creating an animated facial video dataset.