# milliMamba: Specular-Aware Human Pose Estimation via Dual mmWave Radar with Multi-Frame Mamba Fusion

arXiv: https://arxiv.org/abs/2512.20128 · HTML: https://mlsanity.com/paper/2512.20128/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Placed in Video, motion and generative Mamba models, one of 10 directions in State-space sequence models (S4, Mamba).
- Builds on 19 of its 41 references that are in the map.
- First posted on arXiv 23 Dec 2025.

## Placed in

- State-space sequence models (S4, Mamba) / [Video, motion and generative Mamba models](https://mlsanity.com/ssm/video-motion-and-generative-mamba-models/)

## Builds on (most cited first, 19 of 19)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [End-to-End Object Detection with Transformers](https://mlsanity.com/paper/2005.12872/) (2020)
- [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://mlsanity.com/paper/2312.00752/) (2023)
- [Vision Mamba: Efficient Visual Representation Learning with Bidirectional State Space Model](https://mlsanity.com/paper/2401.09417/) (2024)
- [VideoMamba: State Space Model for Efficient Video Understanding](https://mlsanity.com/paper/2403.06977/) (2024)
- [TokenPose: Learning Keypoint Tokens for Human Pose Estimation](https://mlsanity.com/paper/2104.03516/) (2021)
- [SmoothNet: A Plug-and-Play Network for Refining Human Poses in Videos](https://mlsanity.com/paper/2112.13715/) (2021)
- [mm-Pose: Real-Time Human Skeletal Posture Estimation using mmWave Radars and CNNs](https://mlsanity.com/paper/1911.09592/) (2019)
- [HuPR: A Benchmark for Human Pose Estimation Using Millimeter Wave Radar](https://mlsanity.com/paper/2210.12564/) (2022)
- [RT-Pose: A 4D Radar Tensor-based 3D Human Pose Estimation and Localization Benchmark](https://mlsanity.com/paper/2407.13930/) (2024)
- [TCPFormer: Learning Temporal Correlation with Implicit Pose Proxy for 3D Human Pose Estimation](https://mlsanity.com/paper/2501.01770/) (2025)
- [mmPose-NLP: A Natural Language Processing Approach to Precise Skeletal Pose Estimation using mmWave Radars](https://mlsanity.com/paper/2107.10327/) (2021)
- [MMVR: Millimeter-wave Multi-View Radar Dataset and Benchmark for Indoor Perception](https://mlsanity.com/paper/2406.10708/) (2024)
- [GTPT: Group-based Token Pruning Transformer for Efficient Human Pose Estimation](https://mlsanity.com/paper/2407.10756/) (2024)
- [RETR: Multi-View Radar Detection Transformer for Indoor Perception](https://mlsanity.com/paper/2411.10293/) (2024)
- [ProbRadarM3F: mmWave Radar based Human Skeletal Pose Estimation with Probability Map Guided Multi-Format Feature Fusion](https://mlsanity.com/paper/2405.05164/) (2024)
- [Optimizing Human Pose Estimation Through Focused Human and Joint Regions](https://mlsanity.com/paper/2501.14439/) (2025)
- [One Snapshot is All You Need: A Generalized Method for mmWave Signal Generation](https://mlsanity.com/paper/2503.21122/) (2025)

## Built on by (most cited first, 2 of 2)

- [Wave2Body: Rethinking mmWave Human Pose Estimation as Radar-to-Body Token Translation](https://mlsanity.com/paper/2607.18875/) (2026)
- [AViS-Mamba: Adaptive Visual Steering of Audio State-Space Dynamics for Violence Detection](https://mlsanity.com/paper/2604.03329/) (2026)

## Abstract (arXiv)

Millimeter-wave radar offers a privacy-preserving and lighting-invariant alternative to RGB sensors for Human Pose Estimation (HPE) task. However, the radar signals are often sparse due to specular reflection, making the extraction of robust features from radar signals highly challenging. To address this, we present milliMamba, a radar-based 2D human pose estimation framework that jointly models spatio-temporal dependencies across both the feature extraction and decoding stages. Specifically, given the high dimensionality of radar inputs, we adopt a Cross-View Fusion Mamba encoder to efficiently extract spatio-temporal features from longer sequences with linear complexity. A Spatio-Temporal-Cross Attention decoder then predicts joint coordinates across multiple frames. Together, this spatio-temporal modeling pipeline enables the model to leverage contextual cues from neighboring frames and joints to infer missing joints caused by specular reflections. To reinforce motion smoothness, we incorporate a velocity loss alongside the standard keypoint loss during training. Experiments on the TransHuPR and HuPR datasets demonstrate that our method achieves significant performance improvements, exceeding the baselines by 11.0 AP and 14.6 AP, respectively, while maintaining reasonable complexity. Code: https://github.com/NYCU-MAPL/milliMamba