ZEGGS

We present ZeroEGGS, a neural network framework for speech-driven gesture generation with zero-shot style control by example. This means style can be controlled via only a short example motion clip, even for motion styles unseen during training. Our model uses a Variational framework to learn a style embedding, making it easy to modify style through latent space manipulation or blending and scaling of style embeddings. The probabilistic nature of our framework further enables the generation of a variety of outputs given the same input, addressing the stochastic nature of gesture motion. In a series of experiments, we first demonstrate the flexibility and generalizability of our model to new speakers and styles. In a user study, we then show that our model outperforms previous state-of-the-art techniques in naturalness of motion, appropriateness for speech, and style portrayal. Finally, we release a high-quality dataset of full-body gesture motion including fingers, with speech, spanning across 19 different styles.

Publication

Here is a link to the manuscript.

Code

The code, pre-trained models, and our dataset are hosted on GitHub which can be found here.

Referencing ZEGGS

@article{ghorbani2022zeroeggs,
  author = {Ghorbani, Saeed and Ferstl, Ylva and Holden, Daniel and Troje, Nikolaus F. and Carbonneau, Marc-André},
  title = {ZeroEGGS: Zero-shot Example-based Gesture Generation from Speech},
  journal = {Computer Graphics Forum},
  volume = {42},
  number = {1},
  pages = {206-216},
  keywords = {animation, gestures, character control, motion capture},
  doi = {https://doi.org/10.1111/cgf.14734},
  url = {https://onlinelibrary.wiley.com/doi/abs/10.1111/cgf.14734},
  eprint = {https://onlinelibrary.wiley.com/doi/pdf/10.1111/cgf.14734},
  year = {2023}
}