Mem-MLP: Real-Time 3D Human Motion Generation from Sparse Inputs

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Mutlu, Sinan, Angelis, Georgios F., Ozkan, Savas, Wisbey, Paul, Drosou, Anastasios, Ozay, Mete
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917095872135168
author Mutlu, Sinan
Angelis, Georgios F.
Ozkan, Savas
Wisbey, Paul
Drosou, Anastasios
Ozay, Mete
author_facet Mutlu, Sinan
Angelis, Georgios F.
Ozkan, Savas
Wisbey, Paul
Drosou, Anastasios
Ozay, Mete
contents Realistic and smooth full-body tracking is crucial for immersive AR/VR applications. Existing systems primarily track head and hands via Head Mounted Devices (HMDs) and controllers, making the 3D full-body reconstruction in-complete. One potential approach is to generate the full-body motions from sparse inputs collected from limited sensors using a Neural Network (NN) model. In this paper, we propose a novel method based on a multi-layer perceptron (MLP) backbone that is enhanced with residual connections and a novel NN-component called Memory-Block. In particular, Memory-Block represents missing sensor data with trainable code-vectors, which are combined with the sparse signals from previous time instances to improve the temporal consistency. Furthermore, we formulate our solution as a multi-task learning problem, allowing our MLP-backbone to learn robust representations that boost accuracy. Our experiments show that our method outperforms state-of-the-art baselines by substantially reducing prediction errors. Moreover, it achieves 72 FPS on mobile HMDs that ultimately improves the accuracy-running time tradeoff.
format Preprint
id arxiv_https___arxiv_org_abs_2511_16264
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Mem-MLP: Real-Time 3D Human Motion Generation from Sparse Inputs
Mutlu, Sinan
Angelis, Georgios F.
Ozkan, Savas
Wisbey, Paul
Drosou, Anastasios
Ozay, Mete
Computer Vision and Pattern Recognition
Realistic and smooth full-body tracking is crucial for immersive AR/VR applications. Existing systems primarily track head and hands via Head Mounted Devices (HMDs) and controllers, making the 3D full-body reconstruction in-complete. One potential approach is to generate the full-body motions from sparse inputs collected from limited sensors using a Neural Network (NN) model. In this paper, we propose a novel method based on a multi-layer perceptron (MLP) backbone that is enhanced with residual connections and a novel NN-component called Memory-Block. In particular, Memory-Block represents missing sensor data with trainable code-vectors, which are combined with the sparse signals from previous time instances to improve the temporal consistency. Furthermore, we formulate our solution as a multi-task learning problem, allowing our MLP-backbone to learn robust representations that boost accuracy. Our experiments show that our method outperforms state-of-the-art baselines by substantially reducing prediction errors. Moreover, it achieves 72 FPS on mobile HMDs that ultimately improves the accuracy-running time tradeoff.
title Mem-MLP: Real-Time 3D Human Motion Generation from Sparse Inputs
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2511.16264