Audio2Face: Generating Speech/Face Animation from Single Audio with Attention-Based Bidirectional LSTM Networks
This addresses the need for automated facial animation in applications like virtual avatars or video conferencing, representing an incremental improvement with a novel method for a known bottleneck.
The paper tackles the problem of generating real-time facial animation from audio using an end-to-end deep learning approach, achieving accurate lip movements and successful regression of time-varying facial movements.
We propose an end to end deep learning approach for generating real-time facial animation from just audio. Specifically, our deep architecture employs deep bidirectional long short-term memory network and attention mechanism to discover the latent representations of time-varying contextual information within the speech and recognize the significance of different information contributed to certain face status. Therefore, our model is able to drive different levels of facial movements at inference and automatically keep up with the corresponding pitch and latent speaking style in the input audio, with no assumption or further human intervention. Evaluation results show that our method could not only generate accurate lip movements from audio, but also successfully regress the speaker's time-varying facial movements.