VisageSynTalk: Unseen Speaker Video-to-Speech Synthesis via Speech-Visage Feature Selection

Overview

This repository contains the video demo of ECCV2022 paper titled "Unseen Speaker Video-to-Speech Synthesis via Speech-Visage Feature Selection"

Demo video

Each demo video contains the original speech, the generated speech from previous works, and the generated speech from the proposed method from different speakers. The last demo demonstrates the speech sounds differently when we alter the visage style features. The video demos are also availalbe in here.

• GRID Speaker-dependent [Demo Video]
• GRID Speaker-independent [Demo Video]
• LRW Speaker-independent [Demo Video]
• Altering Visage Styles [Demo Video]