After many researchers observed fruitfulness from the recent diffusion probabilistic model, its effectiveness in image generation is actively studied these days. In this paper, our objective is to evaluate the potential of diffusion probabilistic models for 3D human motion-related tasks. To this end, this paper presents a study of employing diffusion probabilistic models to predict future 3D human motion(s) from the previously observed motion. Based on the Human 3.6M and HumanEva-I datasets, our results show that diffusion probabilistic models are competitive for both single (deterministic) and multiple (stochastic) 3D motion prediction tasks, after finishing a single training process. In addition, we find out that diffusion probabilistic models can offer an attractive compromise, since they can strike the right balance between the likelihood and diversity of the predicted future motions.

Designs of our Transformer-based motion denoiser. Inspired by ST-TR, 2CH-TR, and CSDI, our denoiser processes both spatial and temporal information in series (top) or in parallel (bottom). Here, d and t stand for the dimension of each pose-parameter and time, and TF stands for Transformer. Note that the positional encoding also involves adding a learnable vector that represents a diffusion step k as CSDI paper suggests.

Please cite our paper with below bibtex information.

@inproceedings{DiffPred:2023,

title={Can We Use Diffusion Probabilistic Models for 3D Motion Prediction?},

author={Ahn, Hyemin and Mascaro, Valls Esteve an Lee, Dongheui},

year=2023,

month=May,

booktitle={2023 IEEE International Conference on Robotics and Automation (ICRA)},

month_numeric=5

}