LLAniMAtion: LLAMA Driven Gesture Animation

TL;DR

This paper introduces a novel approach for gesture animation in characters using LLAMA2 language model features extracted from text, outperforming traditional audio-driven methods and enabling gesture generation without audio input.

Contribution

The study demonstrates that LLAMA2 features alone can effectively generate synchronized gestures, surpassing audio-based methods, and explores multimodal integration for improved gesture synthesis.

Findings

LLAMA2 features outperform audio features in gesture generation

Combining audio and text features yields no significant improvement over text alone

LLAMA2 enables gesture synthesis without any audio input

Abstract

Co-speech gesturing is an important modality in conversation, providing context and social cues. In character animation, appropriate and synchronised gestures add realism, and can make interactive agents more engaging. Historically, methods for automatically generating gestures were predominantly audio-driven, exploiting the prosodic and speech-related content that is encoded in the audio signal. In this paper we instead experiment with using LLM features for gesture generation that are extracted from text using LLAMA2. We compare against audio features, and explore combining the two modalities in both objective tests and a user study. Surprisingly, our results show that LLAMA2 features on their own perform significantly better than audio features and that including both modalities yields no significant difference to using LLAMA2 features in isolation. We demonstrate that the LLAMA2 based model can generate both beat and semantic gestures without any audio input, suggesting LLMs can provide rich encodings that are well suited for gesture generation.