Despite the recent progress of audio-driven video generation, existing methods mostly focus on driving facial movements, leading to non-coherent head and body dynamics. Moving forward, it is desirable yet challenging to generate holistic human videos with both accurate lip-sync and delicate co-speech gestures w.r.t. given audio. In this work, we propose AudCast, a generalized audio-driven human video generation framework adopting a cascade Diffusion-Transformers (DiTs) paradigm, which synthesizes holistic human videos based on a reference image and a given audio. 1) Firstly, an audio-conditioned Holistic Human DiT architecture is proposed to directly drive the movements of any human body with vivid gesture dynamics. 2) Then to enhance hand and face details that are well-knownly difficult to handle, a Regional Refinement DiT leverages regional 3D fitting as the bridge to reform the signals, producing the final results. Extensive experiments demonstrate that our framework generates high-fidelity audio-driven holistic human videos with temporal coherence and fine facial and hand details. Resources can be found at https://guanjz20.github.io/projects/AudCast.

Speech-driven 3D facial animation has been a long-standing topic due to the complex geometry and motion modeling as well as difficulties in cross-modality learning. Current studies struggle to synthesize human-like lip motions, as they usually represent the movement of the entire face with a compressed global vector, leading to subtle motion loss and thus over-smoothed movements in the local lip region. To cope with this problem, we propose a new speech-driven 3D facial animation framework RQTalker based on the Region-aware Vector Quantization mechanism. The key insight is to first build a region-aware codebook via a self-reconstruction manner, in which each part of the codebook physically corresponds to a facial region with a clear semantic. Our region-aware codebook divides facial movements into local regions for multiple sub-encodings, reducing information loss from compression and improving local facial motion modeling. In addition, we further propose a spatial-temporal Audio-to-Motion Learning Module to produce movements that are spatially more accurate and temporally consistent. Qualitative and quantitative results demonstrate that our method outperforms state-of-the-art approaches.

Evaluating the reliability of intelligent physical systems against rare safety-critical events poses a huge testing burden for real-world applications. Simulation provides a useful platform to evaluate the extremal risks of these systems before their deployments. Importance Sampling (IS), while proven to be powerful for rare-event simulation, faces challenges in handling these learning-based systems due to their black-box nature that fundamentally undermines its efficiency guarantee, which can lead to under-estimation without diagnostically detected. We propose a framework called Deep Probabilistic Accelerated Evaluation (Deep-PrAE) to design statistically guaranteed IS, by converting black-box samplers that are versatile but could lack guarantees, into one with what we call a relaxed efficiency certificate that allows accurate estimation of bounds on the safety-critical event probability. We present the theory of Deep-PrAE that combines the dominating point concept with rare-event set learning via deep neural network classifiers, and demonstrate its effectiveness in numerical examples including the safety-testing of an intelligent driving algorithm.