Search papers, labs, and topics across Lattice.
The paper introduces 4DEquine, a framework for 4D reconstruction of equines from monocular video by disentangling motion and appearance reconstruction. A spatio-temporal transformer with post-optimization is used for motion, while a feed-forward network reconstructs a 3D Gaussian avatar for appearance. Trained on newly created synthetic datasets (VarenPoser and VarenTex), 4DEquine achieves state-of-the-art performance on real-world datasets.
Reconstructing realistic 4D horses from video is now faster and more accurate thanks to a clever disentangling of motion and appearance.
4D reconstruction of equine family (e.g. horses) from monocular video is important for animal welfare. Previous mainstream 4D animal reconstruction methods require joint optimization of motion and appearance over a whole video, which is time-consuming and sensitive to incomplete observation. In this work, we propose a novel framework called 4DEquine by disentangling the 4D reconstruction problem into two sub-problems: dynamic motion reconstruction and static appearance reconstruction. For motion, we introduce a simple yet effective spatio-temporal transformer with a post-optimization stage to regress smooth and pixel-aligned pose and shape sequences from video. For appearance, we design a novel feed-forward network that reconstructs a high-fidelity, animatable 3D Gaussian avatar from as few as a single image. To assist training, we create a large-scale synthetic motion dataset, VarenPoser, which features high-quality surface motions and diverse camera trajectories, as well as a synthetic appearance dataset, VarenTex, comprising realistic multi-view images generated through multi-view diffusion. While training only on synthetic datasets, 4DEquine achieves state-of-the-art performance on real-world APT36K and AiM datasets, demonstrating the superiority of 4DEquine and our new datasets for both geometry and appearance reconstruction. Comprehensive ablation studies validate the effectiveness of both the motion and appearance reconstruction network. Project page: https://luoxue-star.github.io/4DEquine_Project_Page/.
