mmPred: Radar-based Human Motion Prediction in the Dark

Junqiao Fan, Haocong Rao, Jiarui Zhang, Jianfei Yang, Lihua Xie
School of Electrical and Electronic Engineering, Nanyang Technological University, School of Mechanical and Aerospace Engineering, Nanyang Technological University
Paper arXiv Code


Accepted by AAAI 2026

Visualization of mmPred: Given history context (e.g., history raw radar point clouds), mmPred not only estimate history pose sequence (0.5s), but also forecast multiple possiple future poses (1s).

Abstract

Existing Human Motion Prediction (HMP) methods based on RGB-D cameras are sensitive to lighting conditions and raise privacy concerns, limiting their real-world applications such as firefighting and healthcare. Motivated by the robustness and privacy-preserving nature of millimeter-wave (mmWave) radar, this work introduces radar as a novel sensing modality for HMP, for the first time. Nevertheless, radar signals often suffer from specular reflections and multipath effects, resulting in noisy and temporally inconsistent measurements, such as body-part miss-detection. To address these radar-specific artifacts, we propose mmPred, the first diffusion-based frame- work tailored for radar-based HMP.

mmPred introduces a dual-domain historical motion representation to guide the generation process, combining a Time-domain Pose Refinement (TPR) branch for learning fine-grained details and a Frequency-domain Dominant Motion (FDM) branch for capturing global motion trends and suppressing frame-level incon- sistency. Furthermore, we design a Global Skeleton-relational Transformer (GST) as the diffusion backbone to model global inter-joint cooperation, enabling corrupted joints to dynamically aggregate information from others. Extensive exper- iments show that mmPred achieves state-of-the-art performance, outperforming existing methods by 8.6% on mmBody and 22% on mm-Fi.

Motivation

Interpolate start reference image.

(a) mmPred can predict future poses from history radar point clouds, even under darkness. (b) Motivation of mmPred dual-domain design using t-SNE visualization of predicted historical joint locations and velocities. We compare our frequency-domain prediction with the state-of-the-art (SOTA) pose estimator (Yang et al. 2023). Our method produces more distinguishable velocity patterns across actions than existing methods.

System Architecture

Interpolate start reference image.

System architecture: mmPred employs Dual-Domain History Estimation to extract historical motion representations in both the time (TPR) and frequency domains (FDM). These representations are fused via a feature fusion module to construct the condition embedding C, which guides the diffusion-based future motion prediction performed in the frequency domain.

Huamn Pose Prediction Result

Quantative results

Interpolate start reference image.
Interpolate start reference image.

Qualitative result on mmBody dataset

Interpolate start reference image.

Qualitative result on mm-Fi dataset

Interpolate start reference image.

License

mmPred is released under the CC BY-NC 4.0.

Citation

@article{fan2025mmpred,
      title={mmPred: Radar-based Human Motion Prediction in the Dark},
      author={Fan, Junqiao and Rao, Haocong and Zhang, Jiarui and Yang, Jianfei and Xie, Lihua},
      journal={arXiv preprint arXiv:2512.00345},
      year={2025}
}