aiqbench analysis
This research introduces a novel Transformer Autoencoder architecture designed for cross-structural motion retargeting, addressing the limitations of existing models in zero-shot scenarios with unseen skeletal topologies. By integrating graph-based positional encodings multiplicatively into a learnable flattening framework, the method achieves a 43–47% reduction in global joint position error without requiring paired training data. The approach demonstrates that specialized architectural modifications are critical for enabling unified Transformer models to effectively handle diverse skeletal structures in computer vision tasks.
Retargeting Motions to Diverse Skeletons via Learnable Flattening
arXiv:2609.38578v1 Announce Type: cross Abstract: Cross-structural motion retargeting aims to transfer motion between different skeletal topologies. Despite recent progress, existing state-of-the-art models struggle with reliability in zero-shot settings, i.e. skeletons with different topologies which were unseen during training, and recent Transformer-based attempts have failed to outperform specialized geometric methods. We bridge this gap with a Transformer Autoencoder that learns a topology- and translation-invariant latent space. Our core contribution is a learnable flattening of skeletal graphs that captures both local dependencies and global structure. Unlike the standard transformer architecture, which adds positional information to token content, we integrate graph-based positional encodings multiplicatively, a design choice that follows directly from our flattening formulation. The resulting model handles diverse skeletal topologies within a single unified architecture and trains in a fully unsupervised manner, requiring no paired retargeting data. Ablation studies show, that the graph encodings, multiplicative formulation, and Transformer backbone is critical for the performance. In zero-shot evaluations, our method reduces global joint position error by $43-47\%$ over current benchmarks. A user study ($n = 37$), including expert animators, further ranks our approach highest in motion alignment and physical plausibility ($p < 0.05$). These results demonstrate that our model design is key to making transformer architectures effective for motion retargeting, outperforming existing approaches.
Abstract:Cross-structural motion retargeting aims to transfer motion between different skeletal topologies. Despite recent progress, existing state-of-the-art models struggle with reliability in zero-shot settings, i.e. skeletons with different topologies which were unseen during training, and recent Transformer-based attempts have failed to outperform specialized geometric methods. We bridge this gap with a Transformer Autoencoder that learns a topology- and translation-invariant latent space. Our core contribution is a learnable flattening of skeletal graphs that captures both local dependencies and global structure. Unlike the standard transformer architecture, which adds positional information to token content, we integrate graph-based positional encodings multiplicatively, a design choice that follows directly from our flattening formulation. The resulting model handles diverse skeletal topologies within a single unified architecture and trains in a fully unsupervised manner, requiring no paired retargeting data. Ablation studies show, that the graph encodings, multiplicative formulation, and Transformer backbone is critical for the performance. In zero-shot evaluations, our method reduces global joint position error by $43-47\%$ over current benchmarks. A user study ($n = 37$), including expert animators, further ranks our approach highest in motion alignment and physical plausibility ($p < 0.05$). These results demonstrate that our model design is key to making transformer architectures effective for motion retargeting, outperforming existing approaches.
Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
arXiv:2609.38578 [cs.CV]
(or arXiv:2609.38578v1 [cs.CV] for this version)
https://doi.org/10.48550/arXiv.2609.38578 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 305 words
Related coverage
LD4WAM: Learning Latent Dynamics from Human Videos for World Action Models
arXiv · Aug 25, 2026
A Unified Conditional Flow for Motion Generation, Editing, and Intra-Structural Retargeting
arXiv · Aug 28, 2026
Why Cross-Skeleton Retargeting Is Non-Identifiable: Structural Limits of Generative Motion Models
arXiv · Sep 30, 2026
TERRA: Terrain-Aware Reconstruction, Retargeting and Control for Musculoskeletal Locomotion
arXiv · Oct 1, 2026