DAWN: Dynamic Frame Avatar with Non-autoregressive Diffusion Framework for Talking Head Video Generation

Talking head generation intends to produce vivid and realistic talking head videos from a single portrait and speech audio clip. Although significant progress has been made in diffusion-based talking head generation, almost all methods rely on autoregressive strategies, which suffer from limited context utilization beyond the current generation step, error accumulation, and slower generation speed. To address these challenges, we present DAWN (Dynamic frame Avatar With Non-autoregressive diffusion), a framework that enables all-at-once generation of dynamic-length video sequences. Specifically, it consists of two main components: (1) audio-driven holistic facial dynamics generation in the latent motion space, and (2) audio-driven head pose and blink generation. Extensive experiments demonstrate that our method generates authentic and vivid videos with precise lip motions, and natural pose/blink movements. Additionally, with a high generation speed, DAWN possesses strong extrapolation capabilities, ensuring the stable production of high-quality long videos. Furthermore, we hope that DAWN sparks further exploration of non-autoregressive approaches in diffusion models. Talking head generation aims at synthesizing a realistic and expressive talking head from a given portrait and audio clip, which is garnering growing interest due to its potential applications in virtual meetings, gaming, and film production. For talking head generation, it is essential that the lip motions in the generated video precisely match the accompanying speech, while maintaining high overall visual fidelity (Guo et al., 2021a). Furthermore, natural coordination between head pose, eye blinking, and the rhythm of the audio is also crucial for a convincing output (Liu et al., 2023).