arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2207.01567cs.CVcs.AI
收藏

回到MLP:一个简单的人体运动预测基线

Back to MLP: A Simple Baseline for Human Motion Prediction

Inria(法国国家信息与自动化研究所);Univ. Grenoble Alpes(格勒诺布尔阿尔卑斯大学);CNRS(法国国家科学研究中心);Grenoble INP(格勒诺布尔理工学院);Ecole des Ponts(巴黎桥路学院);Univ Gustave Eiffel(古斯塔夫·埃菲尔大学);Tencent AI Lab(腾讯人工智能实验室);Institut de Robòtica i Informàtica Industrial, CSIC-UPC(西班牙高等科学研究理事会-加泰罗尼亚理工大学机器人与工业信息研究所)

Wen Guo, Yuming Du, Xi Shen, Vincent Lepetit, Xavier Alameda-Pineda, Francesc Moreno-Noguer

摘要:

本文解决了人体运动预测的问题,包括从历史上观察到的序列预测未来的身体构成。最先进的方法提供了良好的结果,但是,它们依赖于任意复杂度的深度学习架构,例如递归神经网络(RNN),Transformers或图卷积网络(GCN),通常需要多个训练阶段和超过200万个参数。在本文中,我们表明,结合一系列标准的做法,如应用离散余弦变换(DCT),预测关节的残余位移和优化速度作为辅助损失,一个轻量级的网络基于多层感知器(MLP)只有14万个参数可以超越最先进的性能。对Human3.6M,AMASS和3DPW数据集的详尽评估表明,我们的方法,名为siMLPe,始终优于所有其他方法。我们希望我们的简单方法可以作为社区的一个强有力的基线,并允许重新思考人体运动预测问题。该代码可在\url{https://github.com/dulucas/siMLPe}上公开获得。

英文摘要:

This paper tackles the problem of human motion prediction, consisting in forecasting future body poses from historically observed sequences. State-of-the-art approaches provide good results, however, they rely on deep learning architectures of arbitrary complexity, such as Recurrent Neural Networks(RNN), Transformers or Graph Convolutional Networks(GCN), typically requiring multiple training stages and more than 2 million parameters. In this paper, we show that, after combining with a series of standard practices, such as applying Discrete Cosine Transform(DCT), predicting residual displacement of joints and optimizing velocity as an auxiliary loss, a light-weight network based on multi-layer perceptrons(MLPs) with only 0.14 million parameters can surpass the state-of-the-art performance. An exhaustive evaluation on the Human3.6M, AMASS, and 3DPW datasets shows that our method, named siMLPe, consistently outperforms all other approaches. We hope that our simple method could serve as a strong baseline for the community and allow re-thinking of the human motion prediction problem. The code is publicly available at \url{https://github.com/dulucas/siMLPe}.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑