Search papers, labs, and topics across Lattice.
The paper introduces FA-LAM, a Focus-Aware Large Avatar Model designed for one-shot animatable Gaussian head creation, which also supports static 3D and dynamic 4D full-head recovery. By analyzing attention mechanisms and addressing issues related to noisy activations and conflicting tasks in reconstruction and animation, the authors implement a symmetric attention regularization strategy and a dual-phase training pipeline. The resulting model achieves high-quality reconstruction of animatable heads, particularly excelling in fine facial details and large viewing angles, while maintaining efficiency in multi-view and streaming scenarios.
FA-LAM achieves unprecedented quality in animatable head reconstruction by disentangling the complex tasks of animation and reconstruction, leading to superior performance in fine facial details.
We propose FA-LAM, a Focus-Aware Large Avatar Model for one-shot animatable Gaussian head creation, while simultaneously enabling static 3D and dynamic 4D full-head recovery. The core of our method lies in a thorough analysis of the attention mechanisms and the entangled reconstruction and animation training pipeline adopted by prior state-of-the-art approaches. Our analysis identifies two main factors that compromise the quality of 3D full-head generation: (1) incorrect and noisy attention activations, and (2) conflicts between the tasks of reconstruction and animation. To address the first issue, we introduce a symmetric and semantic attention regularization strategy that leverages the inherent semantics and structural symmetry of human heads. To disentangle the objectives of reconstruction and animation, we develop a novel dual-phase training pipeline that separates the model's capabilities for large-view hallucination and animation into distinct modules. Moreover, we enhance our model to support multi-view and streaming 4D reconstruction in an efficient and memory-friendly manner through a core autoregressive modification with tailored visibility-aware token fusion. Collectively, these innovations enable FA-LAM to reconstruct animatable Gaussian full heads with superior quality, particularly in fine facial regions and large viewing angles.
