FA-LAM: Focus-Aware Large Avatar Model for One-Shot 4D Animatable Gaussian Head

• URL: http://arxiv.org/abs/2607.20922v1
• Date: Thu, 23 Jul 2026 04:57:22 GMT
• Title: FA-LAM: Focus-Aware Large Avatar Model for One-Shot 4D Animatable Gaussian Head
• Abstract summary: FA-LAM is a Focus-Aware Large Avatar Model for one-shot animatable Gaussian head creation.<n>We develop a novel dual-phase training pipeline that separates the model's capabilities for large-view hallucination and animation.<n>We enhance our model to support multi-view and streaming 4D reconstruction in an efficient and memory-friendly manner.
• Score: 17.360653071975936
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: 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.

Related papers

• ReViV: Reconstructing the Viewer and the View in 4D from Monocular Egocentric Video [51.951128822418774]
  ReViV is a framework for holistic egocentric 4D reconstruction.<n>Powered by a Masked Generative Egocentric Transformer, ReViV operates within a single feed-forward architecture.
  arXiv  Detail & Related papers  (2026-07-20T10:22:53Z)
• FFAvatar: Feed-Forward 4D Head Avatar Reconstruction from Sparse Portrait Images [14.13683759834114]
  We present FFAvatar, a Transformer-based 3D Gaussian framework for fast construction of high-quality and animatable 4D head avatars.<n>We introduce a sparse-to-dense learning paradigm to balance visual fidelity and computational efficiency.<n>Experiments demonstrate that FFAvatar efficiently produces high-fidelity and controllable 4D head avatars.
  arXiv  Detail & Related papers  (2026-06-29T14:21:33Z)
• Hand-4DGS: Feed-Forward 3D Gaussian Splatting for 4D Hand Reconstruction from Egocentric Videos [54.612320509347825]
  We introduce Hand-4DGS, the first feed-forward framework for reconstructing dynamic 4D hands directly from egocentric videos.<n>Our approach incorporates a mesh-guided representation for structural priors and temporal convolutions to model dynamic motion.
  arXiv  Detail & Related papers  (2026-06-17T14:58:37Z)
• GSwap: Realistic Head Swapping with Dynamic Neural Gaussian Field [66.31636070843516]
  GSwap is a consistent and realistic video head-swapping system empowered by dynamic neural Gaussian portrait priors.<n>GSwap surpasses existing methods in multiple aspects, including visual quality, temporal coherence, identity preservation, and 3D consistency.
  arXiv  Detail & Related papers  (2026-03-24T13:11:25Z)
• UniMo: Unifying 2D Video and 3D Human Motion with an Autoregressive Framework [54.337290937468175]
  We propose UniMo, an autoregressive model for joint modeling of 2D human videos and 3D human motions within a unified framework.<n>We show that our method simultaneously generates corresponding videos and motions while performing accurate motion capture.
  arXiv  Detail & Related papers  (2025-12-03T16:03:18Z)
• Puppeteer: Rig and Animate Your 3D Models [105.11046762553121]
  Puppeteer is a comprehensive framework that addresses both automatic rigging and animation for diverse 3D objects.<n>Our system first predicts plausible skeletal structures via an auto-regressive transformer.<n>It then infers skinning weights via an attention-based architecture.
  arXiv  Detail & Related papers  (2025-08-14T17:59:31Z)
• LHM: Large Animatable Human Reconstruction Model from a Single Image in Seconds [21.99354901986186]
  We propose LHM (Large Animatable Human Reconstruction Model) to infer high-fidelity avatars represented as 3D Gaussian splatting in a feed-forward pass.<n>Our model leverages a multimodal transformer architecture to effectively encode the human body positional features and image features with attention mechanism.<n>Our LHM generates plausible animatable human in seconds without post-processing for face and hands, outperforming existing methods in both reconstruction accuracy and generalization ability.
  arXiv  Detail & Related papers  (2025-03-13T17:59:21Z)
• MagicMan: Generative Novel View Synthesis of Humans with 3D-Aware Diffusion and Iterative Refinement [23.707586182294932]
  Existing works in single-image human reconstruction suffer from weak generalizability due to insufficient training data or 3D inconsistencies for a lack of comprehensive multi-view knowledge. We introduce MagicMan, a human-specific multi-view diffusion model designed to generate high-quality novel view images from a single reference image.
  arXiv  Detail & Related papers  (2024-08-26T12:10:52Z)
• EgoGaussian: Dynamic Scene Understanding from Egocentric Video with 3D Gaussian Splatting [95.44545809256473]
  EgoGaussian is a method capable of simultaneously reconstructing 3D scenes and dynamically tracking 3D object motion from RGB egocentric input alone. We show significant improvements in terms of both dynamic object and background reconstruction quality compared to the state-of-the-art.
  arXiv  Detail & Related papers  (2024-06-28T10:39:36Z)
• HeadRecon: High-Fidelity 3D Head Reconstruction from Monocular Video [37.53752896927615]
  We study the reconstruction of high-fidelity 3D head models from arbitrary monocular videos. We propose a prior-guided dynamic implicit neural network to tackle these problems.
  arXiv  Detail & Related papers  (2023-12-14T12:38:56Z)
• InvertAvatar: Incremental GAN Inversion for Generalized Head Avatars [40.10906393484584]
  We propose a novel framework that enhances avatar reconstruction performance using an algorithm designed to increase the fidelity from multiple frames. Our architecture emphasizes pixel-aligned image-to-image translation, mitigating the need to learn correspondences between observation and canonical spaces. The proposed paradigm demonstrates state-of-the-art performance on one-shot and few-shot avatar animation tasks.
  arXiv  Detail & Related papers  (2023-12-03T18:59:15Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.