{"ID":2880550,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2508.13562","arxiv_id":"2508.13562","title":"Learnable SMPLify: A Neural Solution for Optimization-Free Human Pose Inverse Kinematics","abstract":"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.","short_abstract":"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...","url_abs":"https://arxiv.org/abs/2508.13562","url_pdf":"https://arxiv.org/pdf/2508.13562v1","authors":"[\"Yuchen Yang\",\"Linfeng Dong\",\"Wei Wang\",\"Zhihang Zhong\",\"Xiao Sun\"]","published":"2025-08-19T06:53:57Z","proceeding":"cs.CV","tasks":"[\"cs.CV\"]","methods":"[]","has_code":false,"code_links":[{"ID":610684,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2880550,"paper_url":"https://arxiv.org/abs/2508.13562","paper_title":"Learnable SMPLify: A Neural Solution for Optimization-Free Human Pose Inverse Kinematics","repo_url":"https://github.com/Charrrrrlie/Learnable-SMPLify","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}