可學習的SMPLify：一種無需優化的人體姿態逆向運動學神經解決方案

Learnable SMPLify: A Neural Solution for Optimization-Free Human Pose Inverse Kinematics

August 19, 2025

作者: Yuchen Yang, Linfeng Dong, Wei Wang, Zhihang Zhong, Xiao Sun

cs.AI

摘要

在三維人體姿態與形狀估計領域，SMPLify作為一個穩健的基準方法，通過迭代優化解決逆運動學（IK）問題。然而，其高昂的計算成本限制了其實用性。近年來，跨領域的研究表明，用數據驅動的神經網絡替代迭代優化，可以在不犧牲精度的情況下顯著提升運行效率。受此趨勢啟發，我們提出了可學習的SMPLify，這是一個神經框架，它用單次回歸模型取代了SMPLify中的迭代擬合過程。我們的框架設計針對神經IK中的兩個核心挑戰：數據構建與泛化能力。為了實現有效訓練，我們提出了一種時間採樣策略，從序列幀中構建初始化-目標對。為了提升對多樣動作及未見姿態的泛化能力，我們採用了以人為中心的歸一化方案和殘差學習，以縮小解空間。可學習的SMPLify支持序列推理和作為插件進行後處理，以精煉現有的基於圖像的估計器。大量實驗證明，我們的方法確立了其作為一個實用且簡潔的基準：相比SMPLify，其運行速度提升了近200倍，在3DPW和RICH數據集上展現出良好的泛化能力，並且在作為LucidAction的插件工具時，能以模型無關的方式運行。代碼已開源於https://github.com/Charrrrrlie/Learnable-SMPLify。

English

In 3D human pose and shape estimation, SMPLify remains a robust baseline that solves inverse kinematics (IK) through iterative optimization. However, its high computational cost limits its practicality. Recent advances across domains have shown that replacing iterative optimization with data-driven neural networks can achieve significant runtime improvements without sacrificing accuracy. Motivated by this trend, we propose Learnable SMPLify, a neural framework that replaces the iterative fitting process in SMPLify with a single-pass regression model. The design of our framework targets two core challenges in neural IK: data construction and generalization. To enable effective training, we propose a temporal sampling strategy that constructs initialization-target pairs from sequential frames. To improve generalization across diverse motions and unseen poses, we propose a human-centric normalization scheme and residual learning to narrow the solution space. Learnable SMPLify supports both sequential inference and plug-in post-processing to refine existing image-based estimators. Extensive experiments demonstrate that our method establishes itself as a practical and simple baseline: it achieves nearly 200x faster runtime compared to SMPLify, generalizes well to unseen 3DPW and RICH, and operates in a model-agnostic manner when used as a plug-in tool on LucidAction. The code is available at https://github.com/Charrrrrlie/Learnable-SMPLify.