Efficient Emotional Adaptation for Audio-Driven Talking-Head Generation
This work addresses the need for cost-effective and efficient emotion control in virtual human applications, though it is incremental as it builds on existing talking-head models.
The paper tackles the problem of inflexibility and inefficiency in audio-driven talking-head synthesis by proposing the Emotional Adaptation for Audio-driven Talking-head (EAT) method, which transforms emotion-agnostic models into emotion-controllable ones using parameter-efficient adaptations, achieving state-of-the-art performance on benchmarks like LRW and MEAD.
Audio-driven talking-head synthesis is a popular research topic for virtual human-related applications. However, the inflexibility and inefficiency of existing methods, which necessitate expensive end-to-end training to transfer emotions from guidance videos to talking-head predictions, are significant limitations. In this work, we propose the Emotional Adaptation for Audio-driven Talking-head (EAT) method, which transforms emotion-agnostic talking-head models into emotion-controllable ones in a cost-effective and efficient manner through parameter-efficient adaptations. Our approach utilizes a pretrained emotion-agnostic talking-head transformer and introduces three lightweight adaptations (the Deep Emotional Prompts, Emotional Deformation Network, and Emotional Adaptation Module) from different perspectives to enable precise and realistic emotion controls. Our experiments demonstrate that our approach achieves state-of-the-art performance on widely-used benchmarks, including LRW and MEAD. Additionally, our parameter-efficient adaptations exhibit remarkable generalization ability, even in scenarios where emotional training videos are scarce or nonexistent. Project website: https://yuangan.github.io/eat/